Estrogen deficiency results in apoptosis in the frontal cortex of adult female aromatase knockout mice

Molecular and Cellular Neuroscience 41 (2009) 1–7

Contents lists available at ScienceDirect

Molecular and Cellular Neuroscience j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y m c n e

Estrogen deficiency results in apoptosis in the frontal cortex of adult female aromatase knockout mice Rachel A. Hill a, Hui Kheng Chua b, Margaret E.E. Jones a, Evan R. Simpson a, Wah Chin Boon a,b,⁎ a b

Prince Henry's Institute of Medical Research, Clayton, VIC 3168, Australia Howard Florey Institute, Parkville, VIC 3052, Australia

a r t i c l e

i n f o

Article history: Received 30 January 2008 Revised 30 November 2008 Accepted 23 December 2008 Available online 8 January 2009 Keywords: Aromatase Estrogen Apoptosis Frontal cortex Female Knockout mouse Brain

a b s t r a c t The aromatase knockout (ArKO) mouse is completely estrogen deficient. We previously detected apoptosis in the hypothalamus of 1 year-old male ArKO mice. This study shows that 12 week-old female ArKO mice display spontaneous apoptosis of pyramidal neurons in the frontal cortex while wild-type (WT) littermates show no signs of apoptosis. Concomitantly, bcl-2 related anti-apoptotic genes are down-regulated whereas the pro-apoptotic gene TRADD is up-regulated in the female ArKO frontal cortex. This phenotype can be rescued by 3-week replacement of 17β-estradiol. Furthermore, the apoptosis phenotype is exacerbated in 12–15 month-old female ArKO mice, which have 30% less neurons in the frontal cortex and lower brain weights than WT counterparts. These data show that estrogens are essential for the survival of female cortical neurons even in the absence of pathological conditions or external assaults. Our observations also demonstrate the sexually dimorphic susceptibility of neurons to estrogen deficiency. © 2009 Elsevier Inc. All rights reserved.

Introduction The presence of aromatase (the enzyme that converts androgens to estrogens) and estrogen receptor (ER) mRNA and protein expression in several regions of the mammalian brain, including the cortex, suggests a strong functional role for estradiol in the brain (Balthazart et al.,1991; Foidart et al., 1995; Jakab et al., 1994). Indeed several studies have reported on the neuroprotective effects of estradiol including antiapoptotic effects, protection from free radicals, anti-inflammatory effects, regulation of calcium channels and protection via increasing cerebral blood flow (see review Amantea et al., 2005). Such neuroprotective properties of estradiol have been implicated in the treatment of several neurological dysfunctions including cognitive and memory deficits, brain injury and stroke (Garcia-Segura et al., 2001; Wise et al., 2001) and in addition neurodegenerative diseases such as Alzheimer's disease (Tang et al., 1996; Xu et al., 2006) and Parkinson's disease (Cyr et al., 2002). While in most cases, the neuroprotective actions of estradiol are ER mediated, non-genomic actions of estradiol are also beneficial — the unique phenolic A ringed structure of estradiol renders it an

⁎ Corresponding author. Howard Florey Institute, c/o The University of Melbourne, VIC 3010, Australia. Fax: +61 3 9348 1707. E-mail address: wah.chin.boon@florey.edu.au (W.C. Boon). 1044-7431/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.mcn.2008.12.009

effective scavenger of free radicals (Moosmann and Behl, 1999). One mechanism of ER-mediated estrogenic neuroprotection is via regulation of anti- and pro-apoptotic genes. Several in vitro studies have demonstrated the anti-apoptotic effects of 17β-estradiol treatment via down-regulation of pro-apoptotic genes such as BAD (Gollapudi and Oblinger, 1999) and up-regulation of anti-apoptotic genes such as Bcl-2 (Dubal et al., 1999; Zhao et al., 2004). In addition, neuroprotective effects of estradiol upon neurodegenerative disease processes such as Alzheimer's and Parkinson's may be mediated through the regulation of anti- and pro-apoptotic gene expression (Pike, 1999; Morissette et al., 2008). Clinically, the topic of estrogen replacement therapy in postmenopausal women has received much attention, and although the literature has been, for the most part, contradictory, it now seems evident that early in menopause, when neurons are still in a healthy state, estrogen therapy can be effective, however if this brief window of opportunity is missed, estradiol treatment is relatively ineffective (Amantea et al., 2005; Maki, 2005; Maki et al., 2007; Stephens et al., 2006). The use of animal models to effectively replicate problems such as hormone deficiency, neurological diseases, brain injury and stroke, is an effective tool for investigating the beneficial effects of estradiol on the brain. Studies using ovariectomy procedures followed by 17β-estradiol replacement have effectively demonstrated the positive effects that estradiol provides in cognitive performances and memory tasks (Gibbs, 2000; Vaucher et al., 2002).

2

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7

Whereas the in vivo effects of estrogen on the brain were primarily investigated using gonadectomised animals (Fukuda et al., 2000; Miller et al., 1999), these animal models are not ideal as aromatase cytochrome P450 (the enzyme that synthesises estrogens from androgens) is known to be expressed in numerous regions of the brain (Lephart et al., 2001; Wagner and Morrell, 1997) such as hypothalamus, amygdala and hippocampus. In studies using aromatase inhibitors, complete blockage in all sites is uncertain. Therefore the total estrogen deficiency can be achieved by knocking out the aromatase gene. Our laboratory has developed an aromatase knockout (ArKO) mouse model of complete estrogen deficiency by deleting exon IX (Fisher et al., 1998) of the aromatase gene (Cyp19A1). Single point mutations in this region have resulted in total estrogen ablation in humans. Neurologically, female ArKO mice show significantly increased ischemic damage following insult in all brain areas, when compared to WT controls (McCullough et al., 2003). In addition, our laboratory has previously determined that 1 year-old male, but not female ArKO mice display apoptosis in the dopaminergic neurons of the arcuate nucleus and medial preoptic area (Hill et al., 2004), indicating that estradiol is not only neuroprotective upon brain injury, but is actually required for the maintenance and integrity of dopaminergic neurons in the hypothalamic region of male mice. Furthermore, we found that ERα agonist treatment prevented the appearance of apoptosis in the arcuate nucleus, while ERβ agonist treatment was neuroprotective in the MPO region (Hill et al., 2007). These data suggest that the role of estradiol in the regulation of dopaminergic neurons is ER mediated. Another ArKO model was generated by deleting exons 1 and 2 of the Cyp19A1, this female ArKO mouse has been shown to exhibit abnormal sexual behaviours (Bakker et al., 2002), and in addition displays depressive-like behaviours that may not be reversed by estradiol treatment (Dalla et al., 2004), suggesting a developmental role for estradiol in the female brain. Furthermore, when crossed with the APP23 Alzheimer's disease model, the combined ArKO/APP23 mice exhibit earlier onset and increased β-amyloid peptide deposition (Yue et al., 2005), suggesting an essential role for estrogens as neuroprotectants from such insults as ischemia and amyloid deposition. This neuroprotective role for estradiol against neurotoxic insults also holds true in the male ArKO which shows enhanced hippocampal damage following domoic acid induced excitotoxicity when compared to WT (Azcoitia et al., 2001). Herein, we present a comprehensive study on consequences of estrogen deficiency on the morphology of the brain of female mice.

Fig. 1. Histological studies of the female WT and ArKO mice frontal cortex. Immunostaining of neurons in the frontal cortex in 1 year-old female (A) ArKO and (B) WT using a mouse anti-neuronal nuclei (NeuN) monoclonal antibody. Less neurons are present in the ArKO mice. (C) Fluorescent labelling of apoptotic cells by TUNEL staining in the frontal cortex in 1 year-old female ArKO. (D) Absence of TUNEL staining in the frontal cortex of 1 year-old female WT. (E) TUNEL staining in the frontal cortex of 10–12 week-old female ArKO. (F) Estrogen-replacement in 10–12 week-old female ArKO mice dramatically reduced TUNEL staining in the frontal cortex.

replacement, the level of TUNEL staining was greatly reduced (Fig. 1F), indicating that the frontal cortex DNA breaks phenotype could be rescued by 17β-estradiol replacement. However, MRI did not detect any difference between the 1 year-old ArKO and WT mice in volumes of the brain or apparent diffusion coefficients of water (data not shown). Interestingly, we did not detect any apoptosis in the hippocampus, the region associated with learning, in the female ArKO or WT mouse at all ages examined.

Results Immunohistochemistry for active caspase-3 Histological studies The brain tissues from ArKO mice were noticeably more friable than those of WT mice. Preliminary histological analysis by NeuN revealed there were less neurons in the frontal cortex in the 1 year-old female ArKO mouse (Fig. 1A) as compared to the 1 year-old female WT (Fig. 1B). No such differences were noticed in the younger animals (14 week-old). Subsequently, semi-quantitation of NeuN immunostained cells of the frontal cortex (Bregma 3.0 mm to 2.10 mm; Paxinos and Franklin 2004) showed that the 15 month-old female ArKO mouse has significantly less (p b 0.001) neurons in the layerII/III, with an average of 196.8 ± 3.45 neurons/30 μm2 (mean ± SEM) as compared to the WT counterpart (an average of 287.9 ± 3.94 neuron/30 μm2) in the same region. We did not detect such differences in the 12 week-old animals (data not shown). In addition, by using TUNEL staining we detected extensive DNA breaks in the frontal cortex of 1 year-old female ArKO mice (Fig. 1C) but not in the WT counterparts (Fig. 1D). These DNA breaks could also be detected, although less frequently, in the young adult (10–12 week-old) female ArKO mouse frontal cortex (Fig. 1E). After 3 weeks of 17β-estradiol

In order to confirm that DNA strand breaks found by TUNEL labelling were due to the apoptotic process, immunohistochemistry for active caspase-3 (an end stage marker for apoptosis) was performed. While levels of active caspase-3 positive labelling were very low to none in WT female frontal cortex (Fig. 2A), a high expression of active caspase-3 was prominent in the layer II/III pyramidal cells of 1 year-old ArKO female mice (Fig. 2C) but no cells in layer V were labelled. We did not detect any TUNEL or active caspase 3 labelling in the same region in the male ArKO mice (data not shown). Double immunohistochemistry for active caspase-3 and NeuN To determine which cell type was undergoing apoptosis, double immunohistochemistry was performed against active caspase-3 and NeuN (a neuronal antibody). Confocal images shown in Figs. 3A and B demonstrate co-localisation of active caspase-3 with NeuN, indicating that the cell type undergoing apoptosis in the frontal cortex of 1 yearold female ArKO mice are neuronal.

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7

3

Fig. 2. Immunohistochemistry staining for active caspase-3 in frontal cortex female WT and ArKO mice, labelling appears as red cytoplasmic or nucleic staining. (A) 1 year-old (y.o.) WT under a red filter, (B) 1 y.o. WT under a green filter to rule out autofluorescence, (C) 1 y.o. ArKO under a red filter, (D) 1 y.o. ArKO under a green filter to rule out autofluorescence.

RNase protection assays To determine which apoptotic pathway has been activated we proceeded to measure the transcript levels of genes related to apoptosis. RNase protection assays were performed on total RNA isolated from the frontal cortex of 10–12 week old mice. All four bcl-2 related anti-apoptotic genes (see reference (Ranger et al., 2001) for review on bcl-2 related proteins), bfl-1, bcl-W, bcl-xL and bcl-2

Fig. 4. Relative transcript levels in frontal cortex of 10–12 week old female ArKO and WT mice detected by RNase protection assay, n = 5. (A) bfl-1, anti-apoptotic gene; (B) bclW, anti-apoptotic gene; (C) bcl-xL, anti-apoptotic gene; (D) bcl-2, anti-apoptotic gene; (E) TRADD, pro-apoptotic gene; ⁎p b 0.05 compared to WT, analysed by two-way ANOVA. Data expressed as mean ± SEM.

(Figs. 4A–D), examined showed lowered transcript expression levels in the ArKO frontal cortex, especially bfl-1 (p = 0.008) and bcl-W (p = 0.03). Conversely, none of the bcl-2 related pro-apoptotic genes (bak, bax and bad) analysed in this study showed any change in transcript levels (data not shown). Of all the death receptor pathway related genes examined in this study, only TRADD, a pro-apoptotic gene (Hsu et al., 1995), showed a significant increase (p b 0.05) in its transcript levels in the ArKO mouse frontal cortex (Fig. 4E). No differences were detected in the transcript levels of other death receptor associated proteins between the genotypes. Again, 3 weeks of treatment with 17β-estradiol in ArKO mice resulted in reversal of the trends of expression except TRADD expression (Figs. 4A–E). Taking these RNase protection assay data together with TUNEL staining and hypocellularity, it is strongly suggested that apoptosis occurs in the frontal cortex of adult female mice in the absence of estrogens. Brain weights In support of the hypocellularity of the 1 year-old female ArKO brain, upon measurement of brain weight, a trend for a decrease in both wet weight (Fig. 5A, p b 0.1), and dry weight (Fig. 5B, p = 0.06) was observed.

Fig. 3. Double-immunohistochemistry staining for active caspase-3 and NeuN (neuronal antibody) in the frontal cortex of female ArKO mice. Active caspase-3 labelling appears as red and NeuN labelling appears as green. (A) and (B) Colocalisation of active casapse-3 and NeuN in the 1 y.o. female ArKO frontal cortex.

Discussion The results of histological studies (TUNEL and immunostaining), RNase protection assay and 17β-estradiol replacement showed that a

4

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7

Fig. 5. Brain weights of WT and ArKO 1 y.o. female mice. (A) Wet weight, (B) dry weight. n = 10 per group.

consequence of estrogen deficiency in female ArKO mice is apoptosis and cell loss in the frontal cortex of the brain. Our results demonstrated that estrogen likely exerts its survival effect on the neurons through up-regulation of anti-apoptotic genes (bfl-1, bcl-W, bcl-xL and bcl-2) and down-regulation of at least one pro-apoptotic gene TRADD. It is interesting to note that TRADD is involved in the TNFα induced apoptosis whereas Bfl-1 is a direct transcriptional target of NF-κB which has been shown to block TNFα-induced apoptosis (Zong et al., 1999). Although there have been reports that cultured neurons undergo apoptosis in the absence of estrogen (Belcredito et al., 2000; Honda et al., 2001; Kajta et al., 2001), this is the first in vivo study to show that neurons undergo spontaneous apoptosis in the absence of estrogen; in contrast to cell death caused by neurotoxin (Azcoitia et al., 2001; Hosoda et al., 2001), ischaemia (Dubal et al., 2001; Fukuda et al., 2000) or impact-accelerated head injury (Roof and Hall, 2000). Regional neuronal hypocellularity in the brain cortex (Wang et al., 2001) has been reported in the estrogen receptor β knockout mouse (ERβKO) but no apoptosis in the brain of these mice has been reported to date. We will not expect ovariectomy of WT animals to result in spontaneous neuronal apoptosis as the aromatase is expressed locally in the brain and the brain of a gonadectomised animal can thus continue to produce estrogens from cholesterol locally (Hojo et al., 2004). It has been previously suggested that not all the neurons that were labelled by TUNEL staining would undergo apoptosis during mouse central nervous system (CNS) development (Gilmore et al., 2000), thus, not all cells labelled by TUNEL staining in the ArKO frontal cortex would undergo apoptosis as the DNA breaks maybe repaired. Nonetheless, no DNA breaks were observed at all in the WT frontal cortex and we were able to rescue this ArKO phenotype by estrogen replacement. Hence, the phenomenon of DNA breaks in the female adult ArKO brain is a consequence of lack of estrogen. Furthermore, the results of the TUNEL are supported by the detection of active caspase 3 in the frontal cortex of the female ArKO and not in the female WT animals. The view that activation of effector caspases such as caspase-3 will eventually lead to apoptosis of a cell may not apply in all cases —

activation of caspase-3 does not necessarily commit a cell to apoptosis. For example, Reddien et al. (2001) reported that in C. elegans, cells expressing CED-3 caspase (homologue of mammalian caspase-3) can survive and even differentiate. Cheng and Zochodne (2003) have also demonstrated that sensory neurons in the dorsal root ganglia of longterm streptozotocin-induced diabetic rats had elevated expression of activated caspase-3 but did not undergo apoptosis. Therefore, the extensive staining of active caspase in layer II/III may not indicate that all the stained neurons will undergo apoptosis at the same time. However, the significantly lower neuronal density of the layer II/III in aged female ArKO mice indicates that a significant proportion of the caspase 3 positive neurons do undergo apoptosis. The fate of the cells depends on the balance of pro- and anti-apoptotic cascades and their interaction with adjacent supporting cells. The observation of significant neuronal loss from the frontal cortex of 12–15 month-old female ArKO mice but not from the 12–14 weekold animals, demonstrated that estrogen is required for maintaining the survival of layer II/III neurons in the female frontal cortex neurons during adult life. In addition, we can also infer that the neuronal loss in the frontal cortex of ArKO mice is not a developmental deficit. The consequences of neuronal loss in the layer II/III of the frontal cortex of female ArKO mice as shown in the current study may cause memory and cognition dysfunction. Several studies have demonstrated negative effects of ovariectomy on the spine density of pyramidal neurons of the frontal cortex in conjunction with reduced performance in memory tasks. In such cases, estradiol replacement was highly effective to increase spine density and consequently improve memory tasks (Wallace et al., 2006). When comparing our data with these previous reports, active caspase-3 labelled cells in aged female ArKO mice, as shown in Fig. 4, do resemble pyramidal cells of the frontal cortex. Consequently, the loss of pyramidal neurons in the frontal cortex may have an impact on memory tasks in female ArKO mice. Previously, it has been reported that the female ArKO mice have compromised performance in the Y-maze test (Martin et al., 2003) but normal performance in the Morris Watermaze test (Boon et al., 2005). Since the latter test is especially sensitive to hippocampal damage, (Morris et al., 1982), the normal performance of female ArKO may reflect the absence of cell death in the female ArKO hippocampus. An important question that arises is whether these observations have any relevance to humans. Temporary retention of verbal or visual information and its active manipulation are intrinsically involved in working memory tasks. The importance of the frontal cortex in working memory has been demonstrated using a wide variety of techniques, including lesion studies in monkeys and patients (Petrides, 1995, 2000) as well as functional neuroimaging in healthy human volunteers (Owen, 1997; Postle et al., 2000). In addition, it has been reported that age-related degeneration of the frontal cortex is greater than the degeneration of other areas of the human brain (Raz et al., 1997) and AD patients had less total prefrontal cortex gray matter than age-matched healthy subjects (Salat et al., 2001). In one study, specifically examining the effect of estrogen on prefrontal cortex-dependent working memory, Duff and Hampson (2000) reported that healthy postmenopausal women taking estrogen exhibited significantly better performance on both verbal (Digit Ordering) and spatial (Spatial Working Memory task) working memory tasks, but did not differ from healthy nonusers on control tasks involving simple passive recall. These results are consistent with a more recent study (Fluck et al., 2002) on the cognitive effects of 10 years of hormone therapy with Tibolone, a drug that contains estrogenic, progestogenic and androgenic activities, on women aged between 54 and 66. Results from this study revealed that women taking Tibolone (when compared to placebo) felt significantly less clumsy. After exposure to a mildly stressful test, the control group felt more anxious, and the treatment group scored significantly higher on semantic memory tests. Carlson and Sherwin (1998) also found that lifetime hormone

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7

therapy use was associated with better baseline modified MiniMental State Examination scores and slower rates of decline. However, data from the Women's Health Initiative have shown that hormone therapy with estrogen and progestin did not improve the cognitive performances of women older than 65 years of age but may increase risk of dementia (Rapp et al., 2003; Shumaker et al., 2003), which illustrates the timing of the administration of the treatment may be of importance for the neuroprotective effects of estrogen (Craig et al., 2005). Interestingly, recent clinical trials have shown that estrogen therapy specifically enhances pre-frontal cortex-dependent cognition processes (Joffe et al., 2006; Krug et al., 2006). In addition, using MRI imaging, a more pronounced activation of the prefrontal cortex was noted in patients receiving hormone therapy including estradiol (Joffe et al., 2006; Smith et al., 2006). These studies suggest a strong role for estradiol in maintaining the integrity of the prefrontal cortex for regulation of memory and cognition, and thus concur with our current data in the mouse. Another interesting observation of this study is the sexually dimorphic susceptibility of neurons to estrogen deficiency. In the male ArKO, apoptosis of dopaminergic neurons in the hypothalamus was detected but not in the frontal cortex (Hill et al., 2004); whereas in the female ArKO, apoptosis of pyramidal neurons in the frontal cortex was observed but no apoptosis was detected in the hypothalamus. These observations may offer clues to the gender bias of Alzheimer's and Parkinson's diseases. In summary, the studies described here provide important new in vivo data that estrogen is essential in maintaining the integrity of the brain frontal cortex in addition to its neuroprotective function against neurotoxins or ischaemia. We show that estrogen preserves the neurons by up-regulating the expression of bcl-2 related antiapoptotic genes. These observations may have significant implications in terms of hormone replacement therapy in postmenopausal women.

5

Immunohistochemistry After 5 h in Bouin's fixative, whole brains were processed into paraffin and sagittal serial sections obtained at 5 μm for histological analysis. Neurons were identified using a mouse anti-neuronal nuclei (NeuN) monoclonal antibody (Chemicon International, Temecula CA) at 1:100. To reduce cross-reactivity of the antibody, sections were pretreated with the M.O.M. Immunodetection kit (Vector, Burlingame CA). The secondary antibody, a biotin-rabbit anti-mouse IgG (H+ L) (1:100; Zymed, San Francisco, CA), was used. The Vectastain ABC and VIP peroxidase substrate kits (Vector, Burlingame CA) were used to amplify and detect the signal, respectively. The ratio of neurons between genotypes was determined by counting a sub-sample of neurons in a randomly selected field of view of known area. Apoptosis was detected using a rabbit anti-active caspase-3 primary antibody (1:40, Ab-4, Oncogene Research Products, San Diego, CA), followed by secondary antibody Alexa Fluor 546 goat anti-rabbit IgG (1:200; Molecular Probes, Eugene, OR). Secondary antibody Alexa Fluor 488 goat anti-mouse IgG (1:200; Molecular Probes, Eugene, OR) was used to detect the NeuN immunostaining. Semi-quantitation of neurons Brains of 12 week-old and 15 month-old female ArKO and WT (n = 5 for each group) animals were harvested and frozen in OCT. Serial coronal sections of 20 μm thickness were collected from the forebrain Bregma 3.08 mm to 2.10 mm (Paxinos and Franklin 2004). Using systematic random sampling, 8 sections from each brain were immunostained with NeuN as previously described. Using the Stereo Investigator (MBF Bioscience, Williston, VT), based on the systematic random sampling principle, twelve 30 μm2 images of the layer II/III of the frontal cortex of each stained sections were taken. Finally, the NeuN positive cells of each image were counted using the Image-Pro® Plus program (Media Cybernetics Inc, MD, USA). The data were analysed statistically by Student's T test.

Experimental methods TUNEL staining Animals ArKO mice (129SV/J X C57BL/6J) were generated by disruption of the Cyp19 gene by homologous recombination (Fisher et al., 1998). Homozygous null or wild type offspring were bred by crossing mice heterozygous for the disrupted gene. The pups were genotyped by PCR as described (Robertson et al., 1999). Animals were housed under SPF conditions and had ad libitum access to water and soy-free mouse chow (Glen Forrest Stockfeeders, Western Australia). Mice were killed by CO2 asphyxiation, brain removed and weighed immediately. For histology studies, brains were placed in Bouin's fixative. For RNA isolation, brains were dissected in RNAlater™ (Ambion Inc, TX, USA), snap frozen in liquid nitrogen and stored at −80 °C. All mice were treated according to Institutional policies of animal welfare and ethical use.

Cells undergoing apoptosis were labelled by TUNEL staining using the In Situ Cell Death Detection Kit (Roche, Mannheim, Germany). Briefly, sagittal paraffin brain sections were dewaxed and rehydrated before treatment with pronase E. After rinsing twice with PBS, the TUNEL reaction mixture (consists of terminal deoxynucleotidyl transferase (TdT) and fluorescein-dUTP) was added. The reactions were performed at 37 °C for 2 h in a dark humidified chamber. After rinsing twice with phosphate buffer, sections were embedded in DAKO Fluorescent Mounting Medium (DAKO Corporation, Carpinteria, USA) and analysed under a fluorescence microscope (Olympus BX50, Hamburg, Germany). Images were captured using the AIS imaging program (Imaging Research Inc, Ontario, Canada). RNase protection assay

Measurements of brain weights One year-old female ArKO and WT (10 animals for each group) were killed by CO2 asphyxiation. After removal from the skull, brains were weighed immediately to obtain the wet weight. To obtain the dry weight, brains were snap frozen in dry ice, lyophilised for 5 days and weighed immediately once removed from the vacuum. Estrogen replacement Placebo or 21-day release 17β-estradiol pellets (0.05 mg, Innovative Research of America, Sarasota, Florida) were implanted subcutaneously in 7–9 week old female mice. After 3 weeks, brain tissues were collected as described above.

Total RNA was isolated (UltraspecTM RNA isolation system (Biotecx, Texas, USA) from the frontal cortex which was dissected according to anatomical landmarks such as optic chasm and fornix. RiboQuant™, a multi-Probe RNase protection assay system (PharMingen, San Diego, USA), was used to examine the expression of apoptotic genes. Two cDNA plasmid templates were applied: mAPO-2 containing the bcl-2 family of genes (bcl-2, bfl1, bcl-W, bcl-xL, bak, bax and bad) and mAPO-3 containing the death receptor pathway genes (Caspase-8, FASL, FAS, FADD, FAP, FAF, TRAIL, TNFRp55, TRADD and RIP). Each assay also included 2 housekeeping genes, L32 and GAPDH. Briefly, antisense 32P-riboprobes were transcribed by T7 RNA polymerase from the cDNA plasmid templates and [α-32P]UTP (Amersham Biosciences, England UK) at 37 °C, 2 h. After phenol chloroform

6

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7

extraction, the 32P-riboprobes were ethanol precipitated. The pellet was dissolved and allowed to hybridised overnight to total RNA at 56 °C. The next day, any unbound probe or free single stranded RNA was digested with RNase A and RNase T1. The protected 32Priboprobes were electrophoresed in a denaturing polyacrylamide gel and quantitated by phosphorimaging (Storm, Molecular Dynamics). Acknowledgments This work was supported by the Australian National Health and Medical Research Council, project grant numbers 338510 and 494813. References Amantea, D., Russo, R., Bagetta, G., Corasaniti, M.T., 2005. From clinical evidence to molecular mechanisms underlying neuroprotection afforded by estrogens. Pharmacol. Res. 52, 119–132. Azcoitia, I., Sierra, A., Veiga, S., Honda, S., Harada, N., Garcia-Segura, L.M., 2001. Brain aromatase is neuroprotective. J. Neurobiol. 47, 318–329. Bakker, J., Honda, S.I., Harada, N., Balthazart, J., 2002. The aromatase knock-out mouse provides new evidence that estradiol is required during development in the female for the expression of sociosexual behaviors in adulthood. J. Neurosci. 22, 9104–9112. Balthazart, J., Foidart, A., Surlemont, C., Harada, N., 1991. Distribution of AromataseImmunoreactive Cells in the Mouse Forebrain. Cell Tissue Res. 263, 71–79. Belcredito, S., Brusadelli, A., Maggi, A., 2000. Estrogens, apoptosis and cells of neural origin. J. Neurocytol. 29, 359–365. Boon, W.C., Diepstraten, J., van der, B.J., Jones, M.E., Simpson, E.R., van den, B.M., 2005. Hippocampal NMDA receptor subunit expression and watermaze learning in estrogen deficient female mice. Brain Res. Mol. Brain Res. 140, 127–132. Carlson, L.E., Sherwin, B.B., 1998. Steroid hormones, memory and mood in a healthy elderly population. Psychoneuroendocrinology 23, 583–603. Cheng, C. and Zochodne, D.W., 2003 Sensory neurons with activated caspase-3 survive long-term experimental diabetes 52, 2363–2371. Craig, M.C., Maki, P.M., Murphy, D.G.M., 2005. The women's health initiative memory study: findings and implications for treatment. Lancet Neurol. 4, 190–194. Cyr, M., Calon, F., Morissette, M., Di Paolo, T., 2002. Estrogenic modulation of brain activity: implications for schizophrenia and Parkinson's disease. J. Psychiatry Neurosci. 27, 12–27. Dalla, C., Antoniou, K., Papadopoulou-Daifoti, Z., Balthazart, J., Bakker, J., 2004. Oestrogen-deficient female aromatase knockout (ArKO) mice exhibit depressivelike symptomatology. Eur. J. Neurosci. 20, 217–228. Dubal, D.B., Shughrue, P.J., Wilson, M.E., Merchenthaler, I., Wise, P.M., 1999. Estradiol modulates bcl-2 in cerebral ischemia: a potential role for estrogen receptors. J. Neurosci. 19, 6385–6393. Dubal, D.B., Zhu, H., Yu, J., Rau, S.W., Shughrue, P.J., Merchenthaler, I., Kindy, M.S., Wise, P.M., 2001. Estrogen receptor alpha, not beta, is a critical link in estradiol-mediated protection against brain injury. Proc. Natl. Acad. Sci. U. S. A. 98, 1952–1957. Duff, S.J., Hampson, E., 2000. A beneficial effect of estrogen on working memory in postmenopausal women taking hormone replacement therapy. Horm. Behav. 38, 262–276. Fisher, C.R., Graves, K.H., Parlow, A.F., Simpson, E.R., 1998. Characterization of mice deficient in aromatase (ArKO) because of targeted disruption of the cyp19 gene. Proc. Natl. Acad. Sci. U. S. A. 95, 6965–6970. Fluck, E., File, S.E., Rymer, J., 2002. Cognitive effects of 10 years of hormone-replacement therapy with tibolone. J. Clin. Psychopharmacol. 22, 62–67. Foidart, A., Harada, N., Balthazart, J., 1995. Aromatase-immunoreactive cells are present in mouse brain areas that are known to express high levels of aromatase activity. Cell Tissue Res. 280, 561–574. Fukuda, K., Yao, H., Ibayashi, S., Nakahara, T., Uchimura, H., Fujishima, M., Hall, E.D., 2000. Ovariectomy exacerbates and estrogen replacement attenuates photothrombotic focal ischemic brain injury in rats. Stroke 31, 155–160. Garcia-Segura, L.M., Azcoitia, I., DonCarlos, L.L., 2001. Neuroprotection by estradiol. Prog. Neurobiol. 63, 29–60. Gibbs, R.B., 2000. Long-term treatment with estrogen and progesterone enhances acquisition of a spatial memory task by ovariectomized aged rats. Neurobiol. Aging 21, 107–116. Gilmore, E.C., Nowakowski, R.S., Caviness Jr., V.S., Herrup, K., 2000. Cell birth, cell death, cell diversity and DNA breaks: how do they all fit together? Trends Neurosci. 23, 100–105. Gollapudi, L., Oblinger, M.M., 1999. Estrogen and NGF synergistically protect terminally differentiated, ERalpha-transfected PC12 cells from apoptosis. J. Neurosci. Res. 56, 471–481. Hill, R.A., Pompolo, S., Jones, M.E., Simpson, E.R., Boon, W.C., 2004. Estrogen deficiency leads to apoptosis in dopaminergic neurons in the medial preoptic area and arcuate nucleus of male mice. Mol. Cell. Neurosci. 27, 466–476. Hill, R.A., Chow, J.D.Y., Hillisch, A., Jones, M.E.E., Simpson, E.R., Boon, W.C., 2007. Fas/FasL mediated apoptosis in the arcuate nucleus and medial preoptic area of male ArKO mice is ameliorated by selective Estrogen Receptor alpha and beta agonist treatment respectively. Mol. Cell. Neurosci. 36, 146–157. Honda, K., Shimohama, S., Sawada, H., Kihara, T., Nakamizo, T., Shibasaki, H., Akaike, A., 2001. Nongenomic antiapoptotic signal transduction by estrogen in cultured cortical neurons. J. Neurosci. Res. 64, 466–475.

Hojo, Y., Hattori, T.A., Enami, T., Furukawa, A., Suzuki, K., Ishii, H.T., Mukai, H., Morrison, J.H., Janssen, W.G., Kominami, S., Harada, N., Kimoto, T., Kawato, S., 2004. Adult male rat hippocampus synthesizes estradiol from pregnenolone by cytochromes P45017alpha and P450 aromatase localized in neurons. Proc. Natl. Acad. Sci. U.S.A. 101, 865–870. Hosoda, T., Nakajima, H., Honjo, H., 2001. Estrogen protects neuronal cells from amyloid beta-induced apoptotic cell death. NeuroReport 12, 1965–1970. Hsu, H., Xiong, J., Goeddel, D.V., 1995. The TNF receptor 1-associated protein TRADD signals cell death and NF-kappa B activation. Cell 19 (81), 495–504. Jakab, R.L., Harada, N., Naftolin, F., 1994. Aromatase- (estrogen synthetase) immunoreactive neurons in the rat septal area. A light and electron microscopic study. Brain Res. 664, 85–93. Joffe, H., Hall, J.E., Gruber, S., Sarmiento, I.A., Cohen, L.S., Yurgelun-Todd, D., Martin, K.A., 2006. Estrogen therapy selectively enhances prefrontal cognitive processes: a randomized, double-blind, placebo-controlled study with functional magnetic resonance imaging in perimenopausal and recently postmenopausal women. Menopause 13, 411–422. Kajta, M., Budziszewska, B., Marszal, M., Lason, W., 2001. Effects of 17-beta estradiol and estriol on NMDA-induced toxicity and apoptosis in primary cultures of rat cortical neurons. J. Physiol. Pharmacol. 52, 437–446. Krug, R., Born, J., Rasch, B., 2006. A 3-day estrogen treatment improves prefrontal cortex-dependent cognitive function in postmenopausal women. Psychoneuroendocrinology 31, 965–975. Lephart, E.D., Lund, T.D., Horvath, T.L., 2001. Brain androgen and progesterone metabolizing enzymes: biosynthesis, distribution and function. Brain Res. Brain Res. Rev. 37, 25–37. Maki, P.M., 2005. A systematic review of clinical trials of hormone therapy on cognitive function — effects of age at initiation and progestin use. Future of hormone therapy: what basic science and clinical studies teach us 1052, 182–197. Maki, P.M., Gast, M.J., Vieweg, A.J., Burriss, S.W., Yaffe, K., 2007. Hormone therapy in menopausal women with cognitive complaints — a randomized, double-blind trial. Neurology 69, 1322–1330. Martin, S., Jones, M., Simpson, E., van den Buuse, M., 2003. Impaired spatial reference memory in aromatase-deficient (ArKO) mice. NeuroReport 14, 1979–1982. McCullough, L.D., Blizzard, K., Simpson, E.R., Oz, O.K., Hurn, P.D., 2003. Aromatase cytochrome P450 and extragonadal estrogen play a role in ischemic neuroprotection. J. Neurosci. 23, 8701–8705. Miller, M.M., Hyder, S.M., Assayag, R., Panarella, S.R., Tousignant, P., Franklin, K.B., 1999. Estrogen modulates spontaneous alternation and the cholinergic phenotype in the basal forebrain. Neuroscience 91, 1143–1153. Moosmann, B., Behl, C., 1999. The antioxidant neuroprotective effects of estrogens and phenolic compounds are independent from their estrogenic properties. Proc. Natl. Acad. Sci. U. S. A. 96, 8867–8872. Morissette, M., Al Sweidi, S., Callier, S., Di Paolo, T., 2008. Estrogen and SERM neuroprotection in animal models of Parkinson's disease. Mol. Cell. Endocrinol. 290, 60–69. Morris, R.G., Garrud, P., Rawlins, J.N., O'Keefe, J., 1982. Place navigation impaired in rats with hippocampal lesions. Nature 297, 681–683. Owen, A.M., 1997. The functional organization of working memory processes within human lateral frontal cortex: the contribution of functional neuroimaging. Eur. J. Neurosci 9, 1329–1339. Paxinos, G., Franklin, K.B.J., 2004. The Mouse Brain in Stereotaxic Coordinates: Compact Second. Edition Academic Press. Petrides, M., 1995. Impairments on nonspatial self-ordered and externally ordered working memory tasks after lesions of the mid-dorsal part of the lateral frontal cortex in the monkey. J. Neurosci. 15, 359–375. Petrides, M., 2000. Impairments in working memory after frontal cortical excisions. Adv. Neurol. 84, 111–118. Pike, C.J., 1999. Estrogen modulates neuronal Bcl-xL expression and beta-amyloidinduced apoptosis: relevance to Alzheimer's disease. J. Neurochem. 72, 1552–1563. Postle, B.R., Berger, J.S., Taich, A.M., D'Esposito, M., 2000. Activity in human frontal cortex associated with spatial working memory and saccadic behavior. J. Cogn. Neurosci. 12 (Suppl 2), 2–14. Ranger, A.M., Malynn, B.A., Korsmeyer, S.J., 2001. Mouse models of cell death. Nat. Genet. 28, 113–118. Rapp, S.R., Espeland, M.A., Shumaker, S.A., Henderson, V.W., Brunner, R.L., Manson, J.E., Gass, M.L., Stefanick, M.L., Lane, D.S., Hays, J., Johnson, K.C., Coker, L.H., Dailey, M., Bowen, D., 2003. Effect of estrogen plus progestin on global cognitive function in postmenopausal women: the Women's Health Initiative Memory Study: a randomized controlled trial. JAMA 289, 2663–2672. Raz, N., Gunning, F.M., Head, D., Dupuis, J.H., McQuain, J., Briggs, S.D., Loken, W.J., Thornton, A.E., Acker, J.D., 1997. Selective aging of the human cerebral cortex observed in vivo: differential vulnerability of the prefrontal gray matter. Cereb. Cortex 7, 268–282. Reddien, P.W., Cameron, S., Horvitz, H.R., 2001. Phagocytosis promotes programmed cell death in C. elegans. Nature 412, 198–202. Robertson, K.M., O'Donnell, L., Jones, M.E., Meachem, S.J., Boon, W.C., Fisher, C.R., Graves, K.H., McLachlan, R.I., Simpson, E.R., 1999. Impairment of spermatogenesis in mice lacking a functional aromatase (Cyp19) gene. Proc. Natl. Acad. Sci. U.S.A. 96, 7986–7991. Roof, R.L., Hall, E.D., 2000. Estrogen-related gender difference in survival rate and cortical blood flow after impact-acceleration head injury in rats. J. Neurotrauma 17, 1155–1169. Salat, D.H., Kaye, J.A., Janowsky, J.S., 2001. Selective preservation and degeneration within the prefrontal cortex in aging and Alzheimer disease. Arch. Neurol. 58, 1403–1408. Shumaker, S.A., Legault, C., Rapp, S.R., Thal, L., Wallace, R.B., Ockene, J.K., Hendrix, S.L., Jones III, B.N., Assaf, A.R., Jackson, R.D., Kotchen, J.M., Wassertheil-Smoller, S.,

R.A. Hill et al. / Molecular and Cellular Neuroscience 41 (2009) 1–7 Wactawski-Wende, J., 2003. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women's Health Initiative Memory Study: a randomized controlled trial. JAMA 289, 2651–2662. Smith, Y.R., Love, T., Persad, C.C., Tkaczyk, A., Nichols, T.E., Zubieta, J.K., 2006. Impact of combined estradiol and norethindrone therapy on visuospatial working memory assessed by functional magnetic resonance imaging. J. Clin. Endocrinol. Metabolism 91, 4476–4481. Stephens, C., Bristow, V., Pachana, N.A., 2006. HRT and everyday memory at menopause: a comparison of two samples of mid-aged women. Women Health 43, 37–57. Tang, M.X., Jacobs, D., Stern, Y., Marder, K., Schofield, P., Gurland, B., Andrews, H., Mayeux, R., 1996. Effect of oestrogen during menopause on risk and age at onset of Alzheimer's disease. Lancet 348, 429–432. Vaucher, E., Reymond, I., Najaffe, R., Kar, S., Quirion, R., Miller, M.M., Franklin, K.B., 2002. Estrogen effects on object memory and cholinergic receptors in young and old female mice. Neurobiol. Aging 23, 87–95. Wagner, C.K., Morrell, J.I., 1997. Neuroanatomical distribution of aromatase MRNA in the rat brain: indications of regional regulation. J. Steroid Biochem. Mol. Biol. 61, 307–314. Wallace, M., Luine, V., Arellanos, A., Frankfurt, M., 2006. Ovariectomized rats show decreased recognition memory and spine density in the hippocampus and prefrontal cortex. Brain Res. 1126, 176–182.

7

Wang, L., Andersson, S., Warner, M., Gustafsson, J.A., 2001. Morphological abnormalities in the brains of estrogen receptor beta knockout mice. Proc. Natl. Acad. Sci. U. S. A. 98, 2792–2796. Wise, P.M., Dubal, D.B., Wilson, M.E., Rau, S.W., Bottner, M., Rosewell, K.L., 2001. Estradiol is a protective factor in the adult and aging brain: understanding of mechanisms derived from in vivo and in vitro studies. Brain Res. Brain Res. Rev. 37, 313–319. Xu, H., Wang, R., Zhang, Y.W., Zhang, X., 2006. Estrogen, beta-amyloid metabolism/ trafficking, and Alzheimer's disease. Ann. N. Y. Acad. Sci. 1089, 324–342. Yue, X., Lu, M., Lancaster, T., Cao, P., Honda, S., Staufenbiel, M., Harada, N., Zhong, Z., Shen, Y., Li, R., 2005. Brain estrogen deficiency accelerates A{beta} plaque formation in an Alzheimer's disease animal model. Proc. Natl. Acad. Sci. U. S. A. 102, 19198–19203. Zhao, L., Wu, T.W., Brinton, R.D., 2004. Estrogen receptor subtypes alpha and beta contribute to neuroprotection and increased Bcl-2 expression in primary hippocampal neurons. Brain Res. 1010, 22–34. Zong, W.X., Edelstein, L.C., Chen, C., Bash, J., Gelinas, C., 1999. The prosurvival Bcl-2 homolog Bfl-1/A1 is a direct transcriptional target of NF-kappaB that blocks TNFalpha-induced apoptosis. Genes Dev. 13, 382–387.