| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
in Estradiol-Mediated Protection against Delayed Cell Death
Department of Physiology (D.B.D., S.W.R., M.B.B., S.B.D., P.M.W.), University of Kentucky College of Medicine, Lexington, Kentucky 40536; Department of Neurology (D.B.D.), University of California at San Francisco, San Francisco, California 94143; Department of Psychiatric Medicine (S.W.R.), University of Virginia Health System, Charlottesville, Virginia 22908; Department of Pharmacology (P.J.S.), Merck Research Laboratories, West Point, Pennsylvania 19486; Department of Physiology and Neuroscience (H.Z., J.Y., M.S.K.), Medical College of South Carolina, Charleston, South Carolina 29425; WHRI (P.J.S., I.M.), Wyeth Research, Collegeville, Pennsylvania 19426; Department of Neurobiology, Physiology, and Behavior (A.B.C., S.S., L.M.G., M.B.B., P.M.W.), Division of Biological Sciences, University of California, Davis, Davis, California 95616; Department of Clinical and Experimental Endocrinology (M.B.B.), University of Gottingen, 37075 Gottingen, Germany; and Department of Physiology and Biophysics (P.M.W.), University of Washington, Seattle, Washington 98195
Address all correspondence and requests for reprints to: Phyllis M. Wise, Ph.D., University of Washington, 301 Gerberding Hall, Box 351237, Seattle, Washington 98195-1237. E-mail: pmwise{at}u.washington.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERß, after MCAO and delineated potential roles for each receptor in estradiol-mediated neuroprotection. We quantified cell death in brains at various times after MCAO and analyzed ER expression by RT-PCR, in situ hybridization, and immunohistochemistry. We found that during the first 24 h, the mechanisms of estradiol-induced neuroprotection after MCAO are limited to attenuation of delayed cell death and do not influence immediate cell death. Furthermore, we discovered that ERs exhibit distinctly divergent profiles of expression over the evolution of injury, with ER induction occurring early and ERß modulation occurring later. Finally, we provide evidence for a new and functional role for ER in estradiol-mediated protection of the injured brain. These findings indicate that physiological levels of estradiol protect against delayed cell death after stroke-like injury through mechanisms requiring ER. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have shown that low, physiological concentrations of estradiol are sufficient to exert profound protection against ischemia induced by permanent middle cerebral artery occlusion (MCAO) in the adult (11) and aging (12) brain and have begun to elucidate potential, protective mechanisms. Estrogens act via multiple mechanisms, depending on the dose and type of estrogen administered (13, 14, 15). Our previous studies established that protection by low estradiol levels necessitates pretreatment (11, 16), requires the estrogen receptor (ER)- (17), and involves modulation of gene expression (18, 19, 20). In contrast to physiological levels, pharmacological doses of estradiol may bypass ERs to attenuate oxidative damage (21, 22, 23) and modulate ion channels (24).
In ischemia, compromised regions of brain undergo immediate, necrotic cell death and delayed, programmed cell death (25, 26). Delayed death involves a sequence of events that leads to cellular degeneration, a process that generally requires transcription and activation or inhibition of factors (25, 27). Because cellular suicide can be blocked in ischemia (28), pathways that counteract death signaling hold great promise for novel therapies for neurodegeneration.
Although our understanding of estradiol’s neuroprotective potential has grown considerably, the precise mechanisms of its actions are still under intense investigation. The major goals of this study were to establish whether mechanisms that underlie estradiol-mediated protection involve an attenuation of delayed cell death and to decipher the roles of ERs in this process.
Our results show that physiologic levels of estradiol attenuate delayed cell death through ER-dependent mechanisms of action. These data carry far-reaching implications for the treatment and prevention of neurodegeneration.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
knockout (KO) mice (19–22 g) (29) were similarly ovariectomized (n = 3–5/experimental group) and implanted with a SILASTIC brand capsule (0.062 in/0.125 in, inner/outer diameter, volume 0.035 ml) containing oil or 17ß-estradiol (180 µg/ml) (n = 3–5/experimental group). This paradigm of 17ß-estradiol treatment produces 20 pg/ml in rats (12) and 25 pg/ml (17) in mice. The SILASTIC brand capsules release hormone over time and produce stable levels of 17ß-estradiol in serum (30), and levels are equivalent to basal circulating levels in the estrous cycle of rats (31) and mice (29, 32).
Cerebral ischemia
Seven days after ovariectomy and treatment, rats and mice underwent permanent MCAO and/or sham surgery. Rats were anesthetized with ketamine/acepromazine (80.0/0.52 mg/kg, ip) and then underwent MCAO via previously described methods (11, 12). Briefly, a 4/0 monofilament suture was inserted through the internal carotid artery to the base of the middle cerebral artery. Mice were anesthetized with a mixture of chloral hydrate/xylazine (350.0/4.0 mg/kg, ip) and then underwent permanent MCAO, as previously described (17), using a method modified from Huang et al. (33). Briefly, a 5/0 blue nylon suture was fired at the tip and advanced 11 mm into the internal carotid artery, in which it effectively occluded the vascular territory of the middle cerebral artery. For all rats and mice, body temperature was maintained at normothermia until recovery from anesthesia. Rat brains were collected at 1, 4, 8, 16, and 24 h after the onset of MCAO and 24 h after the onset of sham injury. Mice brains were collected 24 h after the onset of MCAO or sham injury.
Tissue preparation
To determine the temporal evolution of infarct volumes, rats were killed and their brains were collected and frozen, and 16-µm sections were obtained from bregma points +2.2, +0.2, –1.8, and –3.8 mm of each brain. For molecular studies in rats, coronal sections adjacent to bregma +0.2 mm, an area corresponding to the middle of the rostrocaudal extent of the ischemic infarct, were analyzed for ER gene expression by in situ hybridization histochemistry (one section per animal).
To determine the extent of infarct in mice, 16-µm sections were obtained from anterior to posterior areas of mice brains collected at 24 h to determine the extent of injury, as previously described (17). In this study, infarct volumes were measured from a 600-µm coronal region corresponding to the middle of the infarct (bregma +0.06 to –0.74) in animals used in gene expression studies. For molecular studies in mice, coronal sections (three 200-µm sections from each animal) from bregma +0.06 to –0.74, corresponding to the middle of the ischemic infarct, were microdissected and analyzed for ER gene expression by RT-PCR. As previously described (18), the microdissected regions of cortex were anatomically equivalent, included all laminae, and were adjacent to ischemic infarct. Brain sections from rats and mice were either frozen for molecular studies or fixed with paraformaldehyde and stained with hematoxylin and eosin to clearly delineate the extent of ischemic injury (34). The volume of infarct was calculated by integrating the area of injury on stained coronal sections using computer-assisted imaging (National Institutes of Health Image version 1.60) for quantification, as previously described (11, 17). Frozen brain sections adjacent to those stained with hematoxylin and eosin were used for in situ hybridization histochemistry (one section per animal) and PCR studies. The area of the cortex analyzed for gene expression by in situ hybridization or RT-PCR was selected using previously described criteria (18). Briefly, we first examined the extent of injury on stained sections and then analyzed the region apposed to the infarct on adjacent sections. The analyzed cortical regions were anatomically similar while remaining in noninfarcted tissue.
For immunohistochemical studies, after 4, 8, or 16 h of MCAO, rats were anesthetized with ketamine/acepromazine (80.0/0.52 mg/kg, ip) and transcardially perfused with 150 ml of saline (0.9%, ice-cold) followed by 300 ml of a fixative solution (4% paraformaldehyde, 2.5% acrolein), and a subsequent 50–100 ml of ice-cold saline. Brains were removed from the skull and postfixed overnight in the same fixative solution and transferred into ice-cold Tris buffer. NeuroScience Associates (Knoxville, TN) processed the tissue and created 30-µm coronal sections of all the brains. Sections corresponding approximately to Bregma point +0.2 mm (one section per animal), an area corresponding to the middle of the rostrocaudal extent of ischemic infarct, were used for all immunohistochemical studies.
RNA studies
In situ hybridization histochemistry.
To examine the temporal profile and cellular localization of ER and ERß mRNA expression in the ischemic and sham cerebral cortex, we performed in situ hybridization histochemistry as previously described (35, 36). Briefly, 16-µm coronal sections were hybridized with 200 µl of an antisense 35S-labeled riboprobe (6 x 106 disintegrations per minute/probe per slide), 50% formamide hybridization mixture. The mixture contained a cocktail of one unique riboprobe for ER mRNA (ER 800) (35) or two unique riboprobes for ERß mRNA (ERß 285 and ERß 558) (35). The slides were then incubated overnight at 55 C in a humid chamber, then treated with RNase A, and washed at 67 C in 0.1x saline sodium citrate to remove nonspecific labeling. Slides were then dehydrated, apposed to x-ray film for 3 d, and dipped in nitroblue tetrazolium salt-2 nuclear emulsion (Kodak, Rochester, NY). After exposure for 4–8 wk, the slides were photographically processed, stained with cresyl violet, and coverslipped.
Cellular expression of ER and ERß mRNA was analyzed on a Bioquant IV MEG program to quantitate silver grains. A single threshold was set to determine grains vs. background. The level of lighting and contrast was standardized before quantification so that all slides were assessed under equivalent conditions. The perimeter of each labeled cell within a defined 2 mm2 area of parietal cortex was outlined, and the area covered by grains (above threshold) was quantified. The area analyzed was anatomically equivalent in all animals (one section per animal) and corresponded to a region of cortex that included all laminae and was adjacent to ischemic injury, as described previously (18). Cells demonstrating a value five times higher than background were considered labeled.
RT-PCR studies.
To prepare cDNA for PCR studies in mice, total RNA was isolated from microdissected samples by the method of Chomczynski and Sacchi (37). For each sample, obtained as described above, we reverse transcribed 0.5 µg total RNA to produce cDNA in a final reaction volume of 40 µl that contained 2.5 µM random hexamers (PerkinElmer, Branchburg, NJ), 100 U Moloney murine leukemia virus reverse transcriptase (PerkinElmer), 1 mM deoxynucleotide triphosphate mix (Life Technologies, Gaithersburg, MD), 80 U RNAsin (Promega, Madison, WI), 5 mM MgCl2 (Life Technologies), and 1x reaction buffer (Life Technologies). Samples were incubated for reverse transcription at room temperature for 15 min, 37 C for 2 min, 42 C for 1 h, and 99 C for 5 min. To check for genomic contamination, the same procedure was performed on samples in a reaction solution lacking reverse transcriptase.
We used RT-PCR methods that have been well characterized to determine relative changes in mRNA expression (38). We generated standard curves of input RNA and cycle number for ER, ERß, and ß2-microglobulin (ß2-m) to determine the optimum cycle number within the linear range for PCR amplification (data not shown). This was determined to be 33 cycles for both ER and ERß and 26 cycles for ß2-m. These methods have been validated in studies showing that, within the optimal range of amplification, yields of PCR product are linear with respect to input RNA (38).
Stock solutions containing 2.0 mM MgCl2 (for ER and ERß) or 1.5 mM MgCl2 (for ß2-m), 1x reaction buffer, 10 µCi of 32P-dCTP (3000 Ci/mmol; NEN Life Science Products, Boston, MA), 1 µM each primer, and 1.5 U Taq polymerase (Life Technologies) were prepared for the PCRs. For ER and ERß PCR, 1.5 U Taq Ab (Life Technologies) was included in each reaction. The stock solution was aliquoted (49 µl/tube) and one thirtieth of synthesized cDNA (from the reverse transcription reaction) was added to each sample tube. Samples were then thermocycled for PCR amplification (Hybaid; Touchdown Thermocycler, Middlesex, UK). Optimal reaction conditions for ER and ERß were determined to be 30 sec at 93 C, 30 sec at 67 C, and 1 min at 72 C for 33 cycles. For ß2-m, optimal conditions were 30 sec at 93 C, 30 sec at 57 C, and 1 min at 72 C for 26 cycles. After amplification, PCR products were resolved by PAGE. The gels were heat dried, and the products were visualized and quantified using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
The oligonucleotide sequence pairs used for gene amplification of ERs and the control gene generated PCR products of expected sizes that have been sequenced to verify their identities. To examine ER expression, two sets of primers were used. One set, designed to flank the neomycin sequence present in the ERKO mice, was used to confirm the absence of normal ER message in the cortex of ERKO mice: ER (flanking neo sequence) sense primer: 5'-CGGTCTACGGCCAGTCGGGCACC-3' and antisense primer 5'-GTAGAAGGCGGGAGGGCCGGTGTC-3' (39). The second set of ER primers was used for relative quantification of ER expression: ER sense primer: 5'-GTCTGGTCCTGTGAAGGCTGCAA-3' and antisense primer 5'-GCCTTCCAAGTCATCTCTCAGACG-3' (235 bp) (40). It should be noted that the latter primers amplify mutant ER mRNA, even in knockout mice (data not shown), because message is produced but not translated. ß2-M gene expression was determined using sense primer 5'-GCTATCCAGAAAACCCCTCAA-3' and antisense primer 5'-CATGTCTCGATCCCAGTAGACGGT-3' (300 bp) (41).
Protein studies
Single-label immunohistochemistry.
To study ER protein expression, we performed immunohistochemistry (IHC). Brain sections (one per animal) were washed with 0.05 M Tris (pH 7.4) and then incubated for 20 min in 1% (wt/vol) sodium borohydride in Tris buffer to neutralize acrolein in the sections. After a second wash in Tris buffer, sections underwent a 1-h blocking step at room temperature (Tris, 1% Triton X-100, 10% heat inactivated horse serum). Sections were then incubated overnight at room temperature in blocking buffer containing a commercially available polyclonal antibody directed against a peptide representing the last 15 amino acids of rat ER (C1355, 1:20,000; Upstate Cell Signaling Solutions, Lake Placid, NY). The next day, sections were washed again in Tris and incubated for 1 h in blocking buffer containing biotin-conjugated secondary antirabbit IgG (1:1000; Jackson ImmunoResearch, West Grove, PA). After another wash, the sections were incubated for 1 h in avidin-biotin complex solution for (Vectastain kit, Vector Laboratories, Inc., Burlingame, CA). Antibody complexes were visualized with nickel-enhanced 3,3'-diaminobenzidine. In control experiments, either primary antibody (no primary-degree controls) or secondary antibody (no secondary-degree controls) was excluded from the blocking buffer during the appropriate incubation step. ER immunoreactive cells were counted in periinfarct cerebral cortex. As defined for this study, periinfarct cortex is depicted (see Fig. 4A
) and represents the region in which the most dorsomedial extent of infarction can be found in our model of injury, i.e. where the boundary between infarcted and live tissue can be found. The numbers (1, 2, 3, 4; see Fig. 4A) depict subdivisions of periinfarct cortex used in counting ER-immunoreactive cells. These subdivisions are simply a framework imposed as a method of conceptually organizing the cell counts. These regions were identified by landmarks in the microscope field, ensuring consistency between tissue sections. Cell counts from areas 1 and 3 were combined to yield outer periinfarct cortex counts and counts from areas 2 and 4 were combined to yield inner periinfarct cortex counts.
|
protein localization with lectin, a microglial cell marker, glial fibrillary acidic protein (GFAP), an astrocyte cell marker, and NeuN, a neuronal cell marker, we performed double-label fluorescence IHC. The first day of IHC was completed via the single-label protocol described above. Then sections were washed in Tris and incubated for 1 h in blocking buffer containing biotin-conjugated secondary antirabbit IgG (1:1000; Jackson ImmunoResearch). After another wash, sections were incubated for 45 min at room temperature in Tris buffer containing Cy3-conjugated streptavidin (1:1000; Jackson ImmunoResearch). Sections were washed and incubated overnight at room temperature in blocking buffer containing one of three primary antibodies against a cell-type marker NeuN (1:3000; Chemicon, Temecula, CA), GFAP (1:75,000; Sigma), or biotin-labeled lectin from Lycopersicon esculentum (1:100; Sigma). The next day, sections were washed in Tris and incubated for 1 h in blocking buffer containing biotin-conjugated secondary antimouse IgG (1:1000; Jackson ImmunoResearch). Secondary antibody was not required for lectin because the primary lectin antibody is biotin labeled. After another wash, sections were incubated for 45 min at room temperature in Tris buffer containing Cy2-conjugated streptavidin (1:1000; Jackson ImmunoResearch). Sections were mounted onto slides, dried, and coverslipped for imaging on a laser scanning confocal microscope (Leica, Québec, Canada).
Data analysis
All data are expressed as mean ± SE. To determine the evolution of ischemic injury in rats, data were analyzed using complete between-subject ANOVAs. Significant interactions were probed using one-way ANOVAs. To determine whether estradiol and/or injury influenced the temporal expression of ER or ERß, (2 x 5) complete between-subject ANOVAS were performed and probed with two-way ANOVAS. To determine whether gene expression was different in ischemic vs. sham mice, two-way between-subject ANOVAs were performed. Infarct volumes in mice were analyzed using a two-way ANOVA.
To determine whether estradiol and/or injury influenced the number of ER-containing cells over time, data were analyzed by two-way ANOVA. Significant interactions were probed using Newman Keuls tests. All differences were considered significant at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A–F). These representative sections demonstrate that injury evolves over 24 h into an extensive infarct and that the neuroprotective effects of estradiol (Fig. 1F) occur in the late stage of ischemic injury, compared with oil-treated controls (Fig. 1E). We analyzed and quantified the time-dependent evolution of stroke injury in oil- and estradiol-treated rats and found that estradiol fails to exert neuroprotection early after the onset of stroke (1, 4, 8 h), compared with oil-treated controls (Fig. 2A). However, estradiol exerts dramatic protection against delayed brain injury (16, 24 h) (Fig. 2A). Our data clearly demonstrate that the protective effects of estradiol during this 24-h period in stroke injury are strictly limited to attenuation of delayed cell death.
|
|
). Estradiol significantly protects against cortical injury at later time points (16, 24 h) after the onset of stroke (Fig. 2B). The extent of hormone-mediated protection against delayed cell death in the cerebral cortex is amplified when compared with protection of the whole brain (compare Fig. 2A and 2B).
ER and ERß mRNA are differentially modulated after MCAO: estradiol accelerates and amplifies induction of ER
We investigated the temporal profiles of ER and ERß expression in the cerebral cortex after MCAO to determine potential roles for each receptor in neuroprotection. Cellular expression of ER and ERß mRNA in the periinfarct area was quantified using in situ hybridization histochemistry at the various time points in the evolution of ischemia. Figure 3A shows that ER mRNA induction in the cerebral cortex appears within the first 4 h in ischemic injury in both oil- and estradiol-treated animals. In the absence of estradiol, ER mRNA expression peaks later (16 h) in the ischemic injury. In contrast, estradiol significantly accelerates and amplifies ER mRNA induction early in the evolution of brain injury, compared with oil-treated controls. Peak ER mRNA levels are evident at 4 h in estradiol-treated rats; whereas peak mRNA levels are not attained until 12 h later (16 h) in vehicle-treated rats. These results indicate that although ER is not detectable in the uninjured, adult brain, its induction occurs early enough after MCAO to potentially play a functional role in estradiol’s actions against delayed cell death.
|
. Figure 3B shows that in the absence of estradiol, ERß mRNA expression is initially increased, compared with estradiol-treated animals (at 0, 4, and 8 h). However, injury induces a significant decline in ERß in the later stages of ischemia. In the absence of estradiol, ERß decreases by 16 and 24 h, compared with its expression at 0, 4, and 8 h. Estradiol prevents the injury-induced decrease in ERß during the later stages of ischemic injury (at 16 and 24 h), compared with respective oil-treated controls. These results indicate that ERß is present in the uninjured brain and then declines in the later stages of ischemia; estradiol prevents this late decline.
ER protein is increased in neurons of the cerebral cortex after MCAO: estradiol enhances up-regulation
We examined the expression of ER protein at several time points after ischemia to determine whether the up-regulation of mRNA translates into the increased protein. In parallel with our mRNA data, we found that ER protein, measured in the periinfarct cortex (Fig. 4A) is up-regulated in the ischemic cortex of both oil- and estradiol-treated animals. Figure 4B demonstrates that ER protein is present in the ischemic cortex of both ovariectomized, oil-treated and estradiol-treated rats. Figure 4C shows that estradiol enhances this induction by 8 h after the onset of ischemia, compared with oil-treated controls. This estradiol-induced protein enhancement is seen in the outer layers of cortex and occurs several hours after estradiol-induced up-regulation of ER mRNA.
Next, we determined whether the ischemia-induced ER expression is restricted to specific cell types. We found that ER protein in periinfarct, ischemic cortex is expressed in neurons. Figure 5 shows that under confocal microscopy ER is present in NeuN-containing cells and not in GFAP- or lectin-containing cells. These data show that the protein expression of ER in ischemic cortex is exclusive to neurons (NeuN containing cells) and is not found in astrocytes (GFAP containing cells) or microglial (lectin containing) cells at the times examined.
|
function is essential to mechanisms of estradiol-mediated neuroprotection against cell death in estradiol-mediated protection against delayed cell death, we used wild-type and transgenic mice lacking ER and examined ischemic injury and ER expression in the cerebral cortex. Figure 6A is a composite of representative mRNA expression, using RT-PCR, in microdissected regions of the cerebral cortex of oil- and estradiol-treated, ischemic and sham, wild-type and ERKO mice. Neither injury nor estradiol altered the levels of ß2-m mRNA, indicating that the overall number of live cells is represented equally among the experimental groups (data not shown). The expression of ER was normalized to ß2-m for statistical analyses.
|
shows ER induction in the cortex of wild-type mice 24 h after the onset of injury. In parallel with rats, ER is dramatically induced in the cerebral cortex of oil- and estradiol-treated wild-type mice. ER expression in injured samples increased significantly, compared with sham values, and estradiol does not influence the extent of expression at 24 h. Normal ER mRNA is not detected in oil- or estradiol-treated ERKO mice (Fig. 6B); mutant ER mRNA is produced in ERKO mice (data not shown).
Figure 7 shows the extent of injury in oil- and estradiol-treated, wild-type, and ERKO brains that were microdissected for gene expression studies. Consistent with our previous findings (17), estradiol fails to protect in the absence of ER. In wild-type mice, estradiol decreases cortical injury by over 80%, compared with respective oil-treated controls (Fig. 7). In ERKO mice, estradiol fails to exert a protective effect because the extent of cortical injury is equally extensive in ovariectomized, oil- and estradiol-replaced mice (Fig. 7). Collectively, these findings establish that the up-regulation of ER is critical for the mechanisms of estradiol-mediated protection in ischemic brain injury.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERß, are differentially modulated after the evolution of ischemia. Third, our data demonstrate that ER is dramatically increased in neuronal populations of the cerebral cortex. Last, our data extend our previous finding that estradiol fails to protect against cell death in the absence of ER. Previous studies have shown that in response to MCAO, regions of the brain undergo differential patterns of cellular death including an immediate, necrotic cell death, or a delayed, apoptotic cell death (25, 26). Although ischemic neurons undergoing delayed death do not consistently display classic morphological features of apoptosis, numerous studies have shown that apoptotic regulatory factors are pivotal in stroke injury (25, 44, 45). Our data clearly show that estradiol prevents the delayed pattern of cell death that occurs after stroke injury; thus, our data suggest that estradiol protects the brain by suppressing apoptosis. Indeed, an antiapoptotic role for estradiol is supported by reports from our laboratory and others demonstrating estradiol-induced enhancement of survival factors, such as bcl-2 (18, 20, 46, 47, 48, 49), and estradiol suppression of cell death factors (50), such as caspase-3 (51, 52).
Our finding that estrogen receptors are differentially modulated over the evolution of cortical ischemia highlights unique, potential roles for ERs in the brain. The ER subtypes, ER and ERß, are similar in their structure and ability to bind estradiol but are different in their brain distribution (35) and capability of transactivating genes (53). The discovery that at least two ERs exist (54) has led to numerous investigations into their biological roles (55, 56, 57, 58, 59, 60, 61). Our data strongly suggest a role for ERs, specifically for ER, in protection against delayed cell death resulting from neurodegeneration due to ischemia.
ER mRNA is dramatically up-regulated in the cerebral cortex during the early stages of ischemia, and estradiol accelerates and amplifies this early event; the up-regulation of mRNA is followed by increases in ER protein. Thus, our findings establish that ER is induced early enough in injury to account for estradiol’s late actions against cell death and that the receptor is functionally present during estradiol-mediated protection. Our initial discovery that ER is dramatically induced after injury (18) was a complete surprise because the adult cerebral cortex was not thought to express ER. We now appreciate that ER expression is highly plastic throughout the brain, and the phenomenon of ER induction in injury has now been established in several paradigms of neural injury (18, 62, 63). The injury-induced increase in ER is reminiscent of its expression in the developing brain, during a time of extensive differentiation and neurogenesis (64, 65). It is interesting to speculate that reexpression in the adult brain may reflect dedifferentiation, and attempt to reenter the cell cycle, hypothesized to occur in neurodegenerative injury (66).
We found that the induction of ER appears to be limited to neurons of the cerebral cortex during the first 24 h after MCAO in our experimental paradigm and does not occur in astrocytes or microglia. Garcia-Segura and colleagues (63) have shown that ERs appear in astrocytes after longer intervals after amino acid-induced brain injury. It is possible that these ER-positive neurons facilitate and amplify the actions of estradiol through their communications with other neurons and cell types that may not express the receptor. Several studies have examined such interactions (67, 68, 69, 70). Both ER and ERß have been found to colocalize with the IGF-I receptor in neurons of the cerebral cortex. In hippocampal cultures, increased neuronal survival induced by IGF-I is blocked by treatment with antisense oligonucleotide to ER. Together, the results of many studies (69, 71, 72, 73) suggest that estradiol may facilitate the interaction of ER-positive neurons with other neurons and/or astrocytes to allow the release of factors participating in cross-talk essential for neuroprotection.
Whereas ER is a critical link in the mechanisms of protection against cell death, the role of ERß is less clear. Initially, ERß is lower in estradiol-treated animals; we speculate this may reflect down-regulation of the receptor with chronic hormone treatment. Then ischemia-induced changes occur later in the evolution of ischemic injury. Our data demonstrate that after ischemic injury, ERß is down-regulated and estradiol prevents this down-regulation in the late stages of brain injury. Since the discovery of ERß in 1996 (54), many studies have investigated potential roles for this receptor. Studies have suggested a function for ERß in behavior (74), learning and memory (58), neural development (75), feeding (76), and sexual differentiation of the brain (77). We have previously shown that ERß is not essential to neuroprotection in our paradigm because estradiol continues to protect the brain against ischemia in the absence of ERß (17). Because our current data demonstrate delayed changes in injury and estradiol modulation of ERß expression, we speculate that ERß may play a role in events that follow the cessation of cellular death, such as regeneration and neurogenesis. It is interesting to speculate that ERß-dependent signaling may underlie our recent finding that estradiol enhances neurogenesis after MCAO (78).
Our studies have focused attention on the physiological, receptor-mediated mechanisms of estradiol protection. However, it is well known that estradiol may act through numerous and diverse mechanisms, genomic and nongenomic, receptor and nonreceptor mediated, to exert protection in the brain; furthermore, estrogen receptors may act through ligand-independent pathways to induce gene transcription and influence growth factor/neurotransmitter function (79). Generally, physiological levels of estradiol require pretreatment to exert neuroprotection (11, 80, 81), suggesting that at low levels, estradiol’s effects may be mediated genomically through classic intracellular estrogen receptors and that transcription of hormone-responsive genes (82) plays a critical role. In contrast, treatment with pharmacological doses of estradiol, either acutely or even after injury, suggests that estradiol can act through rapid, nongenomic actions, such as N-methyl-D-aspartate receptor modulation and lipid peroxidation reduction to decrease injury in neural tissue (21, 24, 83). Estradiol can also influence second-messenger signaling and thereby induce neuroprotection (51, 84, 85, 86, 87, 88). Together, these studies establish that estradiol may act by multiple mechanisms and that the predominant mechanisms likely depend on factors such as the dose of hormone administered or the nature of injury induced.
In summary, our results clearly establish key mechanisms of estradiol-mediated neuroprotection. Our discovery that estradiol protects against delayed cell death and that ER is a critical player in this process strongly suggests that estradiol may act through ER to inhibit apoptotic signaling. These data carry potential implications for the selective targeting of ERs in the treatment of disease states, particularly in postmenopausal women (7, 8, 9, 89).
|
Footnotes |
|---|
Disclosure statement: All contributing authors have nothing to declare.
First Published Online March 9, 2006
1 D.B.B. and S.W.R. contributed equally to this work. 
Abbreviations: ER, Estrogen receptor; GFAP, glial fibrillary acidic protein; IHC, immunohistochemistry; KO, knockout; ß2-m, ß2-microglobulin; MCAO, middle cerebral artery occlusion.
Received September 14, 2005.
Accepted for publication February 23, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, not -ß, is a critical link in estradiol-mediated protection against brain injury. Proc Natl Acad Sci USA 98:1952–1957 and -ß mRNA in the rat central nervous system. J Comp Neurol 388:507–525[CrossRef][Medline] protein and estrogen responsiveness. Endocrinology 141:2174–2184): protective effects of estrogen on cell survival after serum deprivation. J Neurosci Res 56:99–108[CrossRef][Medline] in the uterus. Proc Nat Acad Sci USA 97:5936–5941 and ß. Mol Biol Cell 15:1262–1272-[125I]iodoestradiol. Endocrinology 126:1112–1124[Abstract]
) point to cells labeled only with antibody to ER
