**3. Estrogen and Alzheimer's disease**

#### **3.1 Introduction**

Alzheimer's disease is characterized as a gradual failure of memory, cognition, and bodily functions, ultimately leading to death. Although the exact etiology and mechanisms are unknown, the abnormal accumulation of a particular protein, called Amyloid β, has long been proposed as the most likely culprit in the pathogenesis of this disease (Hardy & Selkoe, 2002; Tanzi & Bertram, 2005). In a healthy brain, Amyloid β remains at a steady-state level as a result of the metabolic balance between production of Amyloid β from amyloid precursor protein and removal by cellular uptake and proteolytic degradation (Saido, 1998; Selkoe, 2000). Such a dynamic equilibrium, however, could be altered by genetic or environmental factors that may lead to Alzheimer's disease. It has been hypothesized that Amyloid β is folded into a oligomeric form or a fibrillar (cable-like strings) form (Yamin et al., 2008), both of which are more neurotoxic than Amyloid β itself. Of several different Amyloid β peptides produced, products of Amyloid β-40 and Amyloid β-42 residues are the most common constituents of amyloid plaques, and are widely accepted as the primary trigger for Alzheimer's disease (St George-Hyslop, 2000). In brains with early onset Alzheimer's disease, Amyloid β excessively accumulates. This may be due to the mutations of presenelin genes, which provoke the overproduction of Amyloid β from amyloid precursor protein (Hardy, 2004). In late-onset Alzheimer's disease, which constitutes more than 90% of the disease, the excess accumulation of Amyloid β has been associated with abnormal Amyloid β degrading proteases (Nalivaeva et al., 2008).

Women are more likely to develop Alzheimer's disease after adjusting for age (Andersen et al., 1999). After menopause, the decline of estrogen levels in the brain may render neurons more susceptible to age-related neurodegenerative processes (Coffey et al., 1998). Estrogen therapy, when initiated at the onset of menopause, has reduced the risk or delayed the onset of Alzheimer's disease in women (LeBlanc et al., 2001; Zandi et al., 2002). A recent randomized control trial indicated that estrogen treatment had a beneficial effect on verbal memory in men with mild cognitive impairment (Sherwin et al., 2011 in press). However, clinical studies of estrogen therapy in non-demented and menopausal women have yielded inconclusive results (Craig & Murphy, 2010; Sano et al., 2008). In addition, estrogen administration induced beneficial effects on neuronal function and survival through improving mitochondrial function in healthy neurons (Brinton, 2008). When neurons became unhealthy, estrogen exposure had a detrimental effect (Brinton, 2008). This discrepancy may be due to differences in neurological health, age, hormonal status, the severity of symptoms, the type of menopause (surgical vs. natural), and the type of estrogen compound used (Brinton, 2009). Also, the age when estrogen therapy is initiated, may in part determine the

Estrogen and Brain Protection 145

availability through enhancing the uptake of Amyloid β by microglia (R. Li et al., 2000). In vitro estrogen treatment inhibited the formation of toxic Amyloid β oligomers (Morinaga et al., 2007). Finally, estrogen activated Neprilysin, the primary enzyme that degrades Amyloid β, thereby facilitating Amyloid β degradation in human neuroblastoma cells (Liang et al., 2010). It is possible that this effect of estrogen is preceded by estrogen's action on amyloid precursor protein. Several studies support this notion that estrogen treatment profoundly decreased the levels of amyloid precursor protein by enhancing the degradation of this precursor through the α- and β-secretase pathways (Amtul et al., 2010). Alternatively, estrogen may reduce available amyloid precursor protein by stimulating the formation of vesicles that uptake this precursor-protein, thereby precluding maximal generation of Amyloid β (Greenfield et al., 2002). These findings suggest another mechanism underlying estrogen's protection against Alzheimer's disease involving Amyloid β degredation (Liang et al., 2010). Estrogen may also protect the signaling function of protein kinases from Amyloid β. For example, Amyloid β oligomer inhibited the activity of calcium/calmodulindependent protein kinase II and extracellular signal-regulated kinase in a manner ameliorated by estrogen treatment (Logan et al., 2011). In agreement with the protective effect of estrogen on protein kinase, Szego et al. (2011) reported that the function of protein kinases correlated with avoidance learning behavior. In that study, the treatment with Amyloid β oligomers impeded the learning in a manner that was protected by estrogen. These studies suggest a diverse mechanism by which estrogen protects against Amyloid β

The neurotoxicity of Alzheimer's disease is in part mediated by inflammatory processes (McGeer et al., 2006). Glial cells (non neuronal cells) are involved in this process such that Amyloid β activates glial cells to produce pro-inflammatory cytokines like IL-1β, IL-6, and TNF-α. Activated glial cells have the potential to produce large amounts of reactive oxygen species/nitrogen species by various mechanisms (Zhu et al., 2007). Activated astrocytes produced excessive nitric oxide, which reacted with superoxide to form harmful peroxynitrite (Smith et al., 1997). Excess nitric oxide synthetase was also detected in astrocytes surrounding plaques in Alzheimer's disease (Luth et al., 2001). Estrogen interfered with this process by limiting astroglial cells and inhibiting chronic inflammation associated with Alzheimer's disease (Vegeto et al., 2003). The anti-inflammatory effects of estrogen were shown in a primary culture study; estrogen treatment decreased the expression of pro-inflammatory molecules, such as TNF-α and IL-1β, as well as nitric oxide

Vegeto et al. (2006) conducted a study further supporting the protective effects of estrogen on inflammation associated with Alzheimer's disease. They used the APP23 mouse model, a model of Alzheimer's disease that creates chronic neuroinflammation resembling that in Alzheimer's disease. They found that the number of plaques associated with reactive microglia was increased with age (Vegeto et al., 2006). Interestingly, ovariectomy accelerated microglial activation surrounding Amyloid β plaques, whereas estrogen replacement delayed this process. In parallel, they showed that estrogen reduced the expression of inflammatory mediators, such as monocyte chemoattractant protein-1, macrophage inflammatory protein-2, and TNF-α. That study indicates that microglia is a direct target of estrogen action in the brain. All of these findings reinforce the hypothesis

as an attempt to cope with Alzheimer's disease.

synthase and cyclooxygenase-2 in astrocytes (Valles et al., 2010).

**3.4 Neuroinflammation** 

outcome of estrogen therapy and probably estrogen treatment during the peri-menopause has the highest efficacy (Craig & Murphy, 2010; Genazzani et al., 2007).

In diverse animal models of Alzheimer's disease, estrogen has prevented or delayed the development of Alzheimer's disease pathology in particular Amyloid β accumulation and plaque formation (Carroll et al., 2007; Zheng et al., 2002). Mechanistically, estrogen may regulate the production of Amyloid β and in turn, sustain an improved Amyloid β homeostasis by increasing the metabolism of amyloid precursor protein and destabilization of Amyloid β fibrils (Greenfield et al., 2002; Morinaga et al., 2007). Estrogen's bioenergetic protection may also influence Alzheimer's disease. For instance, estrogen prevented the brain from using alternative fuel sources, such as the ketones (Brinton, 2008, 2009). Aromatase catalyzes the conversion of testosterone to estrogen. Not surprisingly, mice lacking aromatase genes (low estrogen production) showed the loss of hippocampal neurons in response to neurotoxins more severely than wild type mice (Azcoitia et al., 2001), suggesting that estrogen spared those neurons. Indeed, the levels of estrogen and aromatase were significantly reduced in the brains of Alzheimer's disease women (Yue et al., 2005). The view of brain estrogen deficiency as a risk factor for developing Alzheimer's disease pathology is consistent with genetic studies showing an association between the aberration of aromatase gene and the risk for Alzheimer's disease (Iivonen et al., 2004). All these studies suggest that estrogen may have the capacity to interfere with the pathways mediating Alzheimer's disease.

#### **3.2 Estrogen synthesis in Alzheimer's disease**

Since estrogen has a potential capacity to control Alzheimer's disease, one therapeutic strategy might be to target the biosynthesis of estrogen. Indeed, numerous studies have tested whether Alzheimer's disease alters the endogenous synthesis of estrogen. While the levels of estrogens were unchanged in the prefrontal cortex of Alzheimer's disease patients (Rosario et al., 2011), the estrogen biosynthetic enzymes such as aromatase and 17βhydroxysteroid dehydrogenase type 1 were upregulated in the late stages of Alzheimer's disease (Luchetti et al., 2011). Studies using immunohistochemistry showed that aromatase expression was upregulated in astrocytes in later stages of Alzheimer's disease (Azcoitia et al., 2003). Another immunochemistry study also detected an increase in the level of aromatase in the hypothalamic neurons of Alzheimer's patients (Ishunina et al., 2005). The increase was especially profound in the Nucleus basalis of Meynert, a nucleus that is strongly affected in Alzheimer's disease (Ishunina et al., 2005). These findings suggest that during Alzheimer's disease, there is an attempt to increase the biosynthesis of estrogen. The aromatase upregulation may be a defense mechanism of brain areas that undergo neurodegeneration. In support of this notion, the reduced levels of testosterone were found in the aging brain of male and female Alzheimer's patients (Rosario et al., 2011; Weill-Engerer et al., 2002). This seems in line with the idea of a compensatory mechanism, since testosterone is used up after it is locally metabolized into neuroprotective estrogen.

#### **3.3 Amyloid β**

Cumulative evidence indicates that estrogen protects against Amyloid β and its toxicity through mechanisms involving Amyloid β degradation and signaling changes. Estrogen deficiency accelerated the formation of Amyloid β plaque in mice (Yue et al., 2005). Estrogen treatment reduced the level of Amyloid β (Jaffe et al., 1994; Xu et al., 1998) and its

outcome of estrogen therapy and probably estrogen treatment during the peri-menopause has

In diverse animal models of Alzheimer's disease, estrogen has prevented or delayed the development of Alzheimer's disease pathology in particular Amyloid β accumulation and plaque formation (Carroll et al., 2007; Zheng et al., 2002). Mechanistically, estrogen may regulate the production of Amyloid β and in turn, sustain an improved Amyloid β homeostasis by increasing the metabolism of amyloid precursor protein and destabilization of Amyloid β fibrils (Greenfield et al., 2002; Morinaga et al., 2007). Estrogen's bioenergetic protection may also influence Alzheimer's disease. For instance, estrogen prevented the brain from using alternative fuel sources, such as the ketones (Brinton, 2008, 2009). Aromatase catalyzes the conversion of testosterone to estrogen. Not surprisingly, mice lacking aromatase genes (low estrogen production) showed the loss of hippocampal neurons in response to neurotoxins more severely than wild type mice (Azcoitia et al., 2001), suggesting that estrogen spared those neurons. Indeed, the levels of estrogen and aromatase were significantly reduced in the brains of Alzheimer's disease women (Yue et al., 2005). The view of brain estrogen deficiency as a risk factor for developing Alzheimer's disease pathology is consistent with genetic studies showing an association between the aberration of aromatase gene and the risk for Alzheimer's disease (Iivonen et al., 2004). All these studies suggest that estrogen may have the capacity to interfere with the pathways

Since estrogen has a potential capacity to control Alzheimer's disease, one therapeutic strategy might be to target the biosynthesis of estrogen. Indeed, numerous studies have tested whether Alzheimer's disease alters the endogenous synthesis of estrogen. While the levels of estrogens were unchanged in the prefrontal cortex of Alzheimer's disease patients (Rosario et al., 2011), the estrogen biosynthetic enzymes such as aromatase and 17βhydroxysteroid dehydrogenase type 1 were upregulated in the late stages of Alzheimer's disease (Luchetti et al., 2011). Studies using immunohistochemistry showed that aromatase expression was upregulated in astrocytes in later stages of Alzheimer's disease (Azcoitia et al., 2003). Another immunochemistry study also detected an increase in the level of aromatase in the hypothalamic neurons of Alzheimer's patients (Ishunina et al., 2005). The increase was especially profound in the Nucleus basalis of Meynert, a nucleus that is strongly affected in Alzheimer's disease (Ishunina et al., 2005). These findings suggest that during Alzheimer's disease, there is an attempt to increase the biosynthesis of estrogen. The aromatase upregulation may be a defense mechanism of brain areas that undergo neurodegeneration. In support of this notion, the reduced levels of testosterone were found in the aging brain of male and female Alzheimer's patients (Rosario et al., 2011; Weill-Engerer et al., 2002). This seems in line with the idea of a compensatory mechanism, since

testosterone is used up after it is locally metabolized into neuroprotective estrogen.

Cumulative evidence indicates that estrogen protects against Amyloid β and its toxicity through mechanisms involving Amyloid β degradation and signaling changes. Estrogen deficiency accelerated the formation of Amyloid β plaque in mice (Yue et al., 2005). Estrogen treatment reduced the level of Amyloid β (Jaffe et al., 1994; Xu et al., 1998) and its

the highest efficacy (Craig & Murphy, 2010; Genazzani et al., 2007).

mediating Alzheimer's disease.

**3.3 Amyloid β**

**3.2 Estrogen synthesis in Alzheimer's disease** 

availability through enhancing the uptake of Amyloid β by microglia (R. Li et al., 2000). In vitro estrogen treatment inhibited the formation of toxic Amyloid β oligomers (Morinaga et al., 2007). Finally, estrogen activated Neprilysin, the primary enzyme that degrades Amyloid β, thereby facilitating Amyloid β degradation in human neuroblastoma cells (Liang et al., 2010). It is possible that this effect of estrogen is preceded by estrogen's action on amyloid precursor protein. Several studies support this notion that estrogen treatment profoundly decreased the levels of amyloid precursor protein by enhancing the degradation of this precursor through the α- and β-secretase pathways (Amtul et al., 2010). Alternatively, estrogen may reduce available amyloid precursor protein by stimulating the formation of vesicles that uptake this precursor-protein, thereby precluding maximal generation of Amyloid β (Greenfield et al., 2002). These findings suggest another mechanism underlying estrogen's protection against Alzheimer's disease involving Amyloid β degredation (Liang et al., 2010). Estrogen may also protect the signaling function of protein kinases from Amyloid β. For example, Amyloid β oligomer inhibited the activity of calcium/calmodulindependent protein kinase II and extracellular signal-regulated kinase in a manner ameliorated by estrogen treatment (Logan et al., 2011). In agreement with the protective effect of estrogen on protein kinase, Szego et al. (2011) reported that the function of protein kinases correlated with avoidance learning behavior. In that study, the treatment with Amyloid β oligomers impeded the learning in a manner that was protected by estrogen. These studies suggest a diverse mechanism by which estrogen protects against Amyloid β as an attempt to cope with Alzheimer's disease.

#### **3.4 Neuroinflammation**

The neurotoxicity of Alzheimer's disease is in part mediated by inflammatory processes (McGeer et al., 2006). Glial cells (non neuronal cells) are involved in this process such that Amyloid β activates glial cells to produce pro-inflammatory cytokines like IL-1β, IL-6, and TNF-α. Activated glial cells have the potential to produce large amounts of reactive oxygen species/nitrogen species by various mechanisms (Zhu et al., 2007). Activated astrocytes produced excessive nitric oxide, which reacted with superoxide to form harmful peroxynitrite (Smith et al., 1997). Excess nitric oxide synthetase was also detected in astrocytes surrounding plaques in Alzheimer's disease (Luth et al., 2001). Estrogen interfered with this process by limiting astroglial cells and inhibiting chronic inflammation associated with Alzheimer's disease (Vegeto et al., 2003). The anti-inflammatory effects of estrogen were shown in a primary culture study; estrogen treatment decreased the expression of pro-inflammatory molecules, such as TNF-α and IL-1β, as well as nitric oxide synthase and cyclooxygenase-2 in astrocytes (Valles et al., 2010).

Vegeto et al. (2006) conducted a study further supporting the protective effects of estrogen on inflammation associated with Alzheimer's disease. They used the APP23 mouse model, a model of Alzheimer's disease that creates chronic neuroinflammation resembling that in Alzheimer's disease. They found that the number of plaques associated with reactive microglia was increased with age (Vegeto et al., 2006). Interestingly, ovariectomy accelerated microglial activation surrounding Amyloid β plaques, whereas estrogen replacement delayed this process. In parallel, they showed that estrogen reduced the expression of inflammatory mediators, such as monocyte chemoattractant protein-1, macrophage inflammatory protein-2, and TNF-α. That study indicates that microglia is a direct target of estrogen action in the brain. All of these findings reinforce the hypothesis

Estrogen and Brain Protection 147

2009) by promoting neurite plasticity (Beyer et al., 2000). These effects are either mediated through a direct action on dopaminergic neurons or interactions with local astroglia (Ivanova et al., 2001, 2002). Alternatively, estrogen may act on genetic levels to modulate dopamine. For instance, estrogen regulates dopamine gene expression by activating transcriptional factors (DonCarlos et al., 2009). Estrogen also exerts non-genomic membrane effects, interaction with neurotransmitter receptors, and ionic channel regulation (Garcia-Segura et al., 2009). These studies suggest that estrogen protects against Parkinson's disease

Dopamine transporters mediate the uptake of dopamine from synapses to presynaptic vesicles, thereby restoring depleted vesicular dopamine levels (Jourdain et al., 2005). Estrogen stimulated dopamine uptake by nerve cells through neuronal dopamine transporter (D'Astous et al., 2004). On the other hand, estrogen decreased astroglial dopamine uptake, increasing the available levels of synaptic dopamine. This allowed more synaptic dopamine to be taken up by neurons. These studies suggest a few important points: first, not only dopamine neurons but also nigrostriatal astroglia contribute to the metabolic processes of dopamine (Karakaya et al., 2007); second, astroglia are implicated in estrogen-transmitted neuroprotection during dopamine neuro-degeneration (Morale et al, 2006), and finally, as the complementary action of estrogen on neurons, astrocyte and microglia may represent a potential pharmacological target for Parkinson's disease

In the process of dopamine being catalyzed by monoamine oxidase, a large amount of reactive oxygen species is produced, resulting in cell death (Hastings et al., 1996; Luo et al., 1998). In addition, dopamine aldehyde generated in the oxidative deamination reaction is 1000-fold more toxic than dopamine (Burke, 2003). Dopamine neurons in Parkinson's disease become vulnerable to oxidative stress (Dexter et al., 1989; Sian et al., 1994) perhaps

The brain has a predominant defense mechanism against superoxide radicals through antioxidant enzymes such as superoxide dismutase. Studies have demonstrated that superoxide dismutase is implicated in dopamine and Parkinson's disease. Mutant mice that over-expressed or lacked superoxide dismutase were more resistant to (Przedborski et al., 1992) or vulnerable to (Andreassen et al., 2001; J. Zhang et al., 2000) dopamine neurotoxin than wild type mice, respectively. The expression of superoxide dismutase was upregulated in the substantia nigra following the dopamine neurotoxin insult, yet the loss of dopaminergic neurons still occurred (Tripanichkul et al., 2007). These results suggest that there is an attempt to combat the oxidative stress in nigral neurons but not sufficient to spare neurons. The implication of superoxide dismutase in the antioxidant effect of estrogen has been shown in a study done by Tripanichkul et al. (2007). In that study, estrogen treatment increased the expression of superoxide dismutase in the substantia nigra of animals that were treated with the dopamine neurotoxin. This study suggests that estrogen up-regulates superoxide dismutase in critical brain areas, thereby exerting protection

Neuroinflammation and microglial activation are often seen in Parkinson's disease (McGeer et al., 1988; Hunot et al., 2003) and anti-inflammatory drugs reduce the risk of this disease

due to lower levels of glutathione (endogenous antioxidant) than other cell types.

through genomic and non-genomic effects on the dopamine system.

management (Vegeto et al., 2008).

against dopamine neurotoxin or Parkinson's disease.

**4.3 Oxidative stress** 

**4.4 Neuroinflammation** 

that inflammatory mechanisms significantly contribute to the pathogenesis of Alzheimer's disease and support the use of estrogen in the fight against Alzheimer's disease.

Collectively, animal studies on Alzheimer's disease have shown beneficial effects of estrogen through inhibiting the synthesis of amyloid β, facilitating its metabolisms, modulating protein kinases, and inhibiting inflammatory pathways. Human studies on the effects of estrogen on Alzheimer's disease have resulted in both positive and negative effects. It is unclear what causes the inconsistent results. Nevertheless, it seems clear that estrogen influences Alzheimer's disease pathology, if not etiology. How to identify and adjust factors underlying the discrepancies seems to be an essential task.
