| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
D. P. Purpura Department of Neuroscience Albert Einstein College of Medicine Bronx, New York 10461
Address all correspondence and requests for reprints to: Dr. Anne M. Etgen, D. P. Purpura Department of Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461. E-mail: Etgen{at}aecom.yu.edu.
Alzheimer’s disease (AD), the most common cause of dementia in the elderly, is more prevalent in women than in men, and menopause, with its associated estrogen deficiency and altered lipid profile, is considered a significant risk factor for AD (1, 2). There is also good clinical evidence that surgical removal of the ovaries before menopause increases the risk of dementia and cognitive impairment in women (3). However, data from the Women’s Health Initiative Memory Study indicate that hormone treatment instituted years after menopause does not protect against dementia or cognitive decline in elderly women (4, 5, 6). Another major risk factor for AD is expression of the apolipoprotein E (apoE) allele apoE 
4 (7), which is associated with altered circulating cholesterol levels. Experiments in animal models suggest that the well-described neurotrophic effects of estrogens depend on apoE, and that the apoE 4 variant is unable to support estrogen-dependent neurite outgrowth (8, 9). Therefore, it is reasonable to propose that the beneficial effects of estrogen treatment on cognitive function in menopausal women may be influenced by the status of the cholesterol biosynthetic and metabolic pathways as well as other aspects of lipid metabolism (see Ref. 10). The paper by Luciani et al. (11) in this issue of Endocrinology provides yet another clue to the possible interactions between cholesterol and estrogens in the etiology of AD. The ability of 17β-estradiol (E2) to attenuate β-amyloid- and oxidative stress-induced toxicity in human fetal neurons was abolished when small interfering RNA (siRNA) methods were used to silence the expression of seladin-1 (selective Alzheimer’s disease indicator-1) mRNA. Seladin-1 encodes the cholesterol-synthesizing enzyme 3β-hydroxysterol-
-24-reductase (also known as DHCR24), a gene whose expression level was found to be selectively reduced in cortical regions that are particularly vulnerable to AD pathology (12, 13). E2 treatment increased cell cholesterol content, and this response was also blocked in neurons treated with seladin-1 siRNA. The authors went on to clone the human seladin-1 promoter and to identify functional, half-palindromic estrogen response elements (EREs) upstream of the transcription start site.
What sets these findings apart from the dozens of other reports documenting neuroprotective effects of E2 and other estrogens in cultured cells and in vivo (see Refs. 10 and 14 for recent reviews)? First, Luciani and colleagues are one of the few investigators studying human-derived neuronal cells (perhaps more correctly termed neuroblasts). There is still considerable controversy regarding the adequacy of transgenic mouse models to mimic AD neuropathology and cognitive impairment. The long-term cell cultures used in this report (11) are derived from human fetal olfactory neuroepithelium and exhibit many properties of mature olfactory neurons, including electrical excitability, expression of neuron-specific proteins, and responsiveness to odorants (15). They also express endogenous estrogen receptor (ER)
and ERβ, obviating the need to transfect cells and allaying concerns about interpreting the responses of cells that overexpress genes of interest. However, it is legitimate to question how readily one can extrapolate findings in long-term cultures of fetal neuroblasts to the age-related neurodegeneration associated with AD. There is already evidence that expression of seladin-1 is down-regulated when human mesenchymal stem cells are differentiated into neurons (16).
Second, the functional analysis of the seladin-1 promoter in this study (11) suggests that the gene is a direct target of transcriptional activation by ER acting at half-palindromic EREs. Transient transfection assays in Chinese hamster ovary cells demonstrated that enhancement of promoter activity required both ligand and ER. The activity of ERβ was not evaluated, presumably because their previous studies showed that diarylpropionitrile, the ERβ agonist, has modest effects on seladin-1 mRNA levels in these cells (17). Moreover, two selective ER modulators used clinically for treatment of breast cancer and osteoporosis, tamoxifen and raloxifene, were also able to activate the seladin-1 promoter construct. At some concentrations, these two agents also reduced β-amyloid toxicity and increased seladin-1 mRNA (17). If future studies implicate seladin-1 as an estrogen-regulated cell survival factor in the aging human brain, this opens the possibility of developing targeted pharmacological therapies to slow the progression of the debilitating cognitive dysfunction and dementia of AD.
Third, the use of siRNA methods in the present report (11) provides strong evidence of a causal relationship between the previously observed E2 enhancement of neuron survival in cells exposed to β-amyloid or oxidative stress and the apparently ER-mediated up-regulation of seladin-1 mRNA (17). Despite intense efforts to delineate the mechanism(s) underlying E2’s neuroprotective actions in vivo and in cultured neurons, we are still in the early stages of identifying the critical molecular players that determine whether hormone-treated neurons will survive exposure to ischemia, excitotoxicity, oxidative stress, β-amyloid, trauma, and other insults. Moreover, it is noteworthy in this study that treating the neuroblasts with E2 also increased cell cholesterol, an effect that was abrogated by seladin-1 siRNA. This provides a suggestive link between estrogen protection and increased cell cholesterol. However, it may seem contrary to the notion that elevated plasma cholesterol is a risk factor for AD and that statins, which reduce cholesterol synthesis by blocking 3-hydroxy-3-methyglutaryl coenzyme A reductase, reduce AD in patients (18, 19).
The resolution of this apparent paradox lies in the distinction between circulating and cell membrane cholesterol (see Refs. 20 and 21). The factors that maintain brain cholesterol homeostasis are still poorly understood, but because the blood-brain-barrier restricts transport of cholesterol from the plasma into the central nervous system, most brain cholesterol is synthesized in situ (see Ref. 22). In 2003, hippocampal content of cholesterol was reported to be modestly but significantly lower in AD than in control brains (20). A year later, Abad-Rodriquez et al. (23) reported that hippocampal membranes from AD patients and rodent hippocampal neurons with modest (e.g. 30%) cholesterol depletion had higher levels of amyloid precursor protein and β-secretase colocalization and increased production of amyloid peptides relative to normal humans or control mice. This led them to propose that loss of neuronal membrane cholesterol may foster β-amyloid accumulation in AD brains. The links among membrane cholesterol, seladin-1, and neuropathology are reinforced by other work in transgenic mice and in human neuroblastoma cells. Brains from seladin-1 knockout mice lack cholesterol and have elevated levels of β-amyloid; overexpression of seladin-1 in SH-SY5Y neuroblastoma cells reduces amyloid precursor processing and β-amyloid levels (24). Acute exposure of SH-SY5Y cells to oxidative stress produces high levels of seladin-1, elevated cholesterol concentrations and increases in cholesterol biosynthesis. Overexpression of seladin-1 confers cholesterol-dependent resistance to oxidative stress, and mutations of seladin-1 that abolish its reductase activity also block its protective effect (25). Moreover, amyloid accumulation is reduced in SH-SY5Y cells with pharmacologically elevated membrane cholesterol and in cells overexpressing seladin-1. The opposite occurs in cells depleted of cholesterol by treatment with seladin-1 inhibitors (26).
Many questions remain. Chief among them is whether reduced expression of seladin-1 in specific cortical regions is really a marker of sporadic AD or whether it simply reflects cell loss in those vulnerable brain areas. A recently published study (27) reported extensive expression profiling of cortical neurons collected with laser capture microdissection from six different brain regions of AD-afflicted patients and healthy aged brains, using Affymetrix microarrays. Expression of seladin-1 was significantly lower in AD vs. healthy aged brains in some (e.g. hippocampus, medial temporal gyrus, posterior cingulate cortex) but not all (e.g. entorhinal cortex) regions associated with AD pathology. An initial attempt to link mutations in the seladin-1 gene to familial Alzheimer’s in an Italian population also failed to find an association between AD and two different point mutations in the gene (28). Other critical questions are how aging and circulating estrogens impact cholesterol content of neuronal membranes, the relationships between plasma cholesterol and brain cholesterol trafficking, and the role of neuronal membrane cholesterol in amyloid processing and the development of the amyloid plaques and neurofibrillary tangles that are hallmarks of AD. It will also be important to clarify whether ER is the exclusive regulator of seladin-1 expression and whether only the half-palindromic EREs are involved. Luciani et al. (11) also found an AP-1 site in the seladin-1 promoter, and this potential enhancer element was not included in the reporter construct analyzed for regulation by ER. It is entirely possible that ERβ, which is widely expressed in the adult human cortex and hippocampus (29), as well as ER and ER/ERβ heterodimers, may either enhance or repress seladin-1 transcription depending on the ligand (e.g. E2, tamoxifen, raloxifene) and cell context (30, 31, 32).
 |
Footnotes |
|---|
Received June 9, 2008.
Accepted for publication June 16, 2008.
|
References |
|---|
 Top References |
|---|
and β in the human cortex and hippocampus during development and adulthood. J Comp Neurol 503:790–802[CrossRef][Medline] transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc Natl Acad Sci USA 96:15161–15166
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
