SEARCH

SEARCH BY CITATION

Keywords:

  • aging;
  • cholesterol;
  • Cyp46;
  • hippocampus;
  • reactive oxygen species;
  • tyrosine kinase B

Abstract

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

J. Neurochem. (2011) 116, 747–755.

Abstract

It is well established that memory formation and retention involve the coordinated flow of information from the post-synaptic site of particular neuronal populations to the nucleus, where short and long-lasting modifications of gene expression occur. With age, mnemonic, motor and sensorial alterations occur, and it is believed that extra failures in the mechanisms used for memory formation and storage are the cause of neurodegenerative pathologies like Alzheimer’s disease. A prime candidate responsible for damage and loss of function during aging is the accumulation of reactive oxygen species, derived from normal oxidative metabolism. However, dysfunction in the aged brain is not paralleled by an increase in neuronal death, indicative that the brain is better suited to fight against the death signals generated from reactive oxygen species than against loss-of-function stimuli. A main aim of this laboratory is to understand how neurons perform and survive in the constitutive stress background represented by aging. In this report, we summarize our recent findings in relation to survival.

Abbreviations used:
BDNF

brain-derived neurotrophic factor

DIV

days in vitro

PI3K

phosphoinositide 3-kinase

ROS

reactive oxygen species

SOD

superoxide dismutase

Trk

tyrosine kinase

TrkRs

Trk receptors

TTX

tetrodotoxin

Until a decade ago, it was thought that aging was accompanied by neuronal loss. However, it was recently demonstrated that, in the absence of specific pathologies, the aged brain contains few dead neurons (Morrison and Hof 1997; Teter and Finch 2004). However, the performance of neurons decays in the aged brain and a number of evidences, in vitro and in vivo, indicate that the reactive oxygen species (ROS) generated during respiration are the main cause for dysfunction.

Under normal physiological conditions, up to 1% of the mitochondrial electron flow leads to the formation of superoxide (O2), the primary free radical derived from mitochondria. Although this partially reduced oxygen species can already attack sulphur bridges in a variety of enzymes, and thus perturb their function, O2 can be rapidly converted into hydrogen peroxide (H2O2), and later in water, by the superoxide dismutase (SOD)1, SOD2 and SOD3. Yet, H2O2 can rapidly react with reduced transition metals originating the highly reactive hydroxyl radical (OH), which is far damaging for cells. In addition, O2 can react with nitric oxide to generate peroxynitrite, which reacts with carbon dioxide, leading to protein damage via nitration of tyrosines.

In all cells, the levels of ROS are controlled by a number of detoxification pathways that use biological antioxidants such as glutathione, α-tocopherol (vitamin E), carotenoids and ascorbic acid. In addition, induction of the heat-shock response by heat-shock proteins is essential for survival under the deleterious consequences of stressors, like free radicals. In presence of stressors, protein missfolding occurs, triggering the activation of different heat-shock proteins, which prevent aggregation and promote refolding, or even divert misfolded proteins to the proteasome for posterior destruction (Whitesell and Lindquist 2005). However, when the levels of ROS exceed the capacity of cells for detoxification they enter into a process called allostatic load (McEwen 1999, 2007). Allostasis is the process of complex adaptations in response to stressors whereby the settings for function are displaced but still within the physiological range. Hence, allostatic load is the condition of wear and tear produced by the repeated activation of allostatic mechanisms. Allostatic load has different layers of severity (i.e. reduced dendritic remodelling to increased production of excitotoxins) and therefore different consequences at the level of cognition and motricity, depending on the brain region affected. Although allostatic load hampers function, neurons do not die thanks to the activation of anti-apoptotic/pro-survival pathways.

In the hippocampus, a region critically involved in cognition, pro-survival activity is mainly derived from the activation of the tyrosine kinase (Trk) receptor TrkB. During development, this receptor is activated by the neurotrophin brain-derived neurotrophic factor (BDNF). At this stage, BDNF/TrkB-stimulated intracellular signalling is critical for neuronal survival during the process of programmed cell death. Later on, BDNF/TrkB is important for morphogenesis and, in adulthood, for establishment and maintenance of synaptic plasticity (Reichardt 2006).

Although binding of BDNF to TrkB receptors elicits various intracellular signalling pathways, survival critically depends on the activation of the mitogen-activated protein kinase/extracellular signal-regulated protein kinase and phosphoinositide 3-kinase (PI3K). Activated PI3K inhibits apoptotsis via Akt phosphorylation. Below, we summarize our work on TrkB activity during hippocampal aging.

Primary hippocampal cultures as model system for certain aspects of brain aging

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

In the last years, cell cultures were widely used as models to study the molecular mechanisms of aging, particularly cultured human fibroblasts and T lymphocytes. They contributed to demonstrate the concept of replicative senescence, which is aging at the cellular level and refers to the fact that normal dividing cells have a finite proliferation capacity in vitro (Hayflick 1998). However, an important question remains regarding whether research in replicative senescence will shed light into the aging of post-mitotic differentiated cells. To address this, we and other labs (Kuroda et al. 1995; Porter et al. 1997; Aksenova et al. 1999; Sodero et al. 2010) have proposed long-term primary cultures of neurons as an appropriate model to study the mechanisms of aging. In these cultures, the survival of mature differentiated neurons decreases according to the ‘Gompertz law’, which states an exponential increase of mortality with age (Aksenova et al. 1999). This law applies for the majority of living organisms that are used as experimental models in gerontological studies. Particularly, we propose the use of primary hippocampal cultures, first, because in this system terminal differentiation (synaptogenesis) occurs at a rather fixed time, between 10–12 days in vitro (DIV), allowing comparisons between cells grown for 10–12 DIV (‘young’ mature) and cells grown for 20–24 DIV (‘old’ mature). Second, because hippocampal neurons in vitro progressively develop typical signs of neuronal aging in situ (Martin et al. 2009; Sodero et al. 2010). Thus, after synaptogenesis neurons gradually accumulate lipofuscin granules (aggregates of oxidized protein and lipids), ROS, acetylated tubulin and hyperphosphorylated Tau. Third, because damage accumulation is paralleled by the activation of canonical anti-stress responses (i.e. pJnk), allowing neurons to outlive for almost two weeks after the initial appearance of stress by-products. Although much of the aging defects found in vivo are not reproduced in vitro, these long-term hippocampal cultures do faithfully reproduce a major constituent of aging: stress accumulation and antistress responses. Because our final goal is to understand brain aging, and consequently the pathologies of the aged brain, the results obtained with hippocampal cultures were compared with the in vivo situation. Some of our recent findings are summarized below.

TrkB activity in the mature brain can occur by BDNF complementary mechanisms

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

During development, neurotrophins are mandatory for survival, differentiation and growth of different neuronal populations (Reichardt 2006). In the mature nervous system, neurotrophins are important for the modulation of neuronal connectivity and activity-dependent plasticity (Conover and Yancopoulos 1997; Blum and Konnerth 2005). Neurotrophins bind and activate Trk receptors (TrkRs) that in turn lead to the activation of multiple intracellular signaling pathways, most notoriously those involving mitogen-activated protein kinases and PI3K (Kaplan and Miller 2000; Reichardt 2006). In the hippocampus, a region of the brain critically involved in learning and memory, the most prominently expressed neurotrophin receptor is TrkB (Tokuyama et al. 1998), whose cognate ligand is BDNF.

Although there is no doubt that BDNF is the main modulator of TrkB activity, new evidences indicate that certain roles mediated by TrkB may occur independently from BDNF. For instance, TrkB conditional knockout mice present clear defects in pre- and post-synaptic morphogenesis in the hippocampus (Luikart et al. 2005), although this is not the case in BDNF-conditional knockout mice (Gorski et al. 2003; Hill et al. 2005). These observations are consistent with the lack of an overt effect on the development of the hippocampus in BDNF knockout animals (Ernfors et al. 1994; Jones et al. 1994). In this direction, we have also observed that hippocampal membranes from 21 day-old BDNF −/− mice or 10 month-old BDNF +/− mice present levels of TrkB phosphorylation comparable to those of wild-type controls (Martin et al. 2008). Furthermore, primary hippocampal cultures prepared from BDNF −/− embryos show high TrkB phosphorylation, not different from that of neuronal cultures prepared from wild-type littermates at the same age in vitro (Martin et al. 2008). It is important to highlight that hippocampal neurons in culture are grown without the need of supplementation of BDNF, or any other neurotrophin (Kaech and Banker 2006). Furthermore, the absolute need of BDNF for all and every TrkB-mediated activity in the adult hippocampus is challenged by reports showing that the levels and activity of BDNF in the hippocampus start to decay in the early adulthood in rats, reaching the lowest levels at the end of the first year of life (Gooney et al. 2004; Shetty et al. 2004; Hattiangady et al. 2005; see however Katoh-Semba et al. 1998). We have demonstrated that the levels of phosphorylated TrkB increase in mouse hippocampus as the animal age, from 21-days old to 20 months old (Martin et al. 2008). Like in vivo, the levels of phosphorylated TrkB also increase after terminal differentiation in primary hippocampal cultures. The increase in TrkB phosphorylation in vitro is independent of the presence of any ligand accumulated in the culture medium, because young neurons incubated with conditioned medium obtained from 25 DIV-cultures do not show increased levels of phospho-TrkB (Martin et al. 2008). All the presented evidences suggest that alternative or complementary mechanisms to BDNF binding have evolved during cellular/neuronal differentiation to guarantee TrkB activity in the adult brain. The best characterized of such mechanisms is the TrkB activation by the lower affinity ligands NT3/NT4 (Yan et al. 1993; Davies et al. 1995), although in vivo studies have confirmed that the lack of BDNF is not always compensated by NT4 (Conover et al. 1995; Stenqvist et al. 2005). Robust TrkB signaling in the mature brain can also occur through the increased expression of TrkB subunits, and/or of its co-receptor p75 (Zaccaro et al. 2001; Hartmann et al. 2004; Marshak et al. 2007). The number of contributions that support a ligand-independent activation of TrkRs is steadily increasing (Paratcha and Ibanez 2002; Suzuki et al. 2004; Nicolau et al. 2006; Hanzal-Bayer and Hancock 2007; Jacobson et al. 2007; Pereira and Chao 2007). It is also possible that TrkB signaling in the adult brain occurs via up-regulation of ligands like adenosine or gangliosides, which have been shown to act as potent inducers of TrkB activity (Lee and Chao 2001; Duchemin et al. 2002).

High TrkB activity can be induced by membrane cholesterol reduction

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

Previous reports showing the occurrence of lipid changes during neuronal aging (Kracun et al. 1992; Valsecchi et al. 1996; Prinetti et al. 2001; Modi et al. 2008; Sugiura et al. 2008; Yamamoto et al. 2008) made us hypothesize that these changes could have been the cause for the high TrkRs activity in the adult brain. As a matter of fact, Trk activity requires clustering into detergent resistant membrane domains (Paratcha and Ibanez 2002; Pereira and Chao 2007). Consistent with our hypothesis, we observed higher TrkB activity and lower cholesterol content in neurons maintained in culture for 25 days compared to young ones. A similar dichotomy (i.e. high active TrkB and low cholesterol) was observed in hippocampal membranes from old mice (Martin et al. 2008). The existence of a direct link between low cholesterol and high TrkB receptor activation was directly demonstrated by pharmacological manipulation of cholesterol in primary hippocampal neurons. Thus, a reduction of 20–30% of membrane cholesterol in young neurons in vitro was sufficient to elicit TrkB activation in a reversible manner (Martin et al. 2008). A similar effect was observed for TrkA in PC12 cells subjected to a similar cholesterol removal–replenishment protocol (Iannilli et al. 2009).

The demonstration that cholesterol reduction can suffice to increase TrkRs activity constitutes, at first sight, a paradigmatic example of signalling control from rafts. In fact, in most of the cases cholesterol loss is associated with reduced signalling (Simons and Toomre 2000). However, our results are supported by recent publications in which it was shown that cholesterol reduction can lead to increased signalling in different cell types, including developing neurons (Suzuki et al. 2004; Kalvodova et al. 2005; Oh et al. 2007). Also, comparing 7 and 21 DIV hippocampal neurons, Nicholson and Ferreira (2009) have reported that higher levels of membrane cholesterol in mature neurons might increase susceptibility to Aβ induced toxicity. Thus, a conclusion from our results is that membrane cholesterol variations can result in different responses, depending on the signalling pathway, the type of cell and the stage of differentiation. Next, we investigated the mechanisms involved in cholesterol loss in the aged hippocampus.

Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

Low-membrane cholesterol in the aged neurons may have occurred through different mechanisms, acting individually or in conjunction, that is, reduced neuronal sinthesis, reduced import from astrocytes or increased removal from neurons (see Dietschy and Turley 2001). Quantitative PCR revealed a significant increase in the expression levels of the cytochrome P450 enzyme cholesterol 24-hydroxylase (Cyp46) in old-mature hippocampal neurons in culture compared with young-mature neurons (Martin et al. 2008). A similar age-associated increase was observed when RNA from the hippocampus of mice of different ages was analyzed. Cyp46 converts cholesterol into 24(S)-hydroxycholesterol (Bjorkhem and Diczfalusy 2004), which diffuses out of cells, crosses the blood–brain barrier, and is cleared by the liver (Lutjohann et al. 1996; Bjorkhem et al. 1997, 1998, 2001; Li-Hawkins et al. 2000). Importantly, high levels of this enzyme are found in neuronal cells, specifically in pyramidal neurons of the hippocampus and cortex, in Purkinje cells of the cerebellum, and in hippocampal and cerebellar interneurons (Lund et al. 1999; Ramirez et al. 2008). In agreement with our findings, Lund et al. (1999) had previously shown that the levels of Cyp46 increase with aging, both in the mouse and human brain, suggesting that in fact cholesterol loss in hippocampal neurons aging in vitro and in vivo can be the consequence of increased Cyp46 amount. Next, we defined the cause behind the up-regulation of Cyp46 in the aged hippocampus.

Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

Ohyama et al. (2006) have reported that the transcription of the cyp46A1 gene, encoding for Cyp46, was insensitive to sterols, which normally regulate HMGCoA reductase and Cyp7A, two key enzymes for cholesterol synthesis and elimination, respectively. Furthermore, and especially relevant in the context of aging, these same authors demonstrated that oxidative stress is a potent inductor of Cyp46 expression (Ohyama et al. 2006). In agreement, cyp46A1 was found induced under conditions accompanied by high stress, such as cortical injury, induced autoimmune encephalomyelitis and Alzheimer’s disease (Bogdanovic et al. 2001; Teunissen et al. 2007; Cartagena et al. 2008). Considering the above, we investigated whether or not metabolic stress was responsible for Cyp46/cholesterol loss in aged neurons.

As mentioned above, in normative aging the brain suffers morphological and functional modifications affecting mainly the synapses, and a growing number of evidences have implicated reactive oxygen species (ROS) as the cause for these changes. Like in other cells, neuronal ROS are generated at many cellular sites, but especially at the mitochondria, during the respiratory chain. In addition, a growing number of evidences suggest that ROS can also be produced in association with Ca2+ bursts derived from NMDA receptor activation (Ha et al. 2010). Interestingly, it has recently been described that most of the ROS produced by an excessive activation of NMDA receptors requires the non-mitochondrial NADPH oxidase enzyme (Brennan et al. 2009). Because NADPH oxidase enzyme is a plasma membrane enzyme (Bedard and Krause 2007), this opens the possibility of significant amounts of ROS would be produced out of the mitochondria.

In agreement with the above, using hippocampal cultures aging in vitro we demonstrated that excitatory neurotransmission is a potent inducer of cholesterol loss (Sodero et al. 2010). In this work, chronic reduction of electrical activity by tetrodotoxin (TTX) administration prevented cholesterol loss while acute NMDA treatment induced it. The observation that cholesterol loss induced by NMDA was precluded by the NMDA receptor antagonist dl-2-amino-5-phosphonopentanoic acid strongly indicates that the effect of TTX was via reducing glutamatergic neurotransmission (Sodero et al. 2010). In addition, in this work we observed that excitatory neurotransmission induced the up-regulation of the cholesterol-hydroxylating enzyme Cyp46, in turn because of accumulation of ROS. In fact, synaptic activity increased Cyp46 at the mRNA and protein levels and this increase was prevented by chronic blockage of voltage-gated sodium channels with TTX.

To determine if the existence of a direct association between stress accumulation, cholesterol decrease and survival activation was an exclusive phenomenon of hippocampal neurons in vitro, experiments were performed in fully differentiated PC12 cells subjected to stress. Nutrient deprivation of already differentiated cells has been utilized as experimental paradigm for studying cellular stress in several cell types, including PC12 cells (Li et al. 2007). Using this system, we confirmed that stress was able to induce generation of ROS, as well as membrane cholesterol loss and TrkA activation (Iannilli et al. 2009). Furthermore, this work confirmed that ROS production via NADPH oxidase was required for cholesterol loss induced by starvation. In addition, increased levels of ROS produced by the direct administration of organic peroxide (tert-butyl-hydroperoxide) to differentiated but non-starved PC12 cells, resulted in cholesterol loss. Although the results in PC12 cells confirm that cholesterol loss may be a common mechanism to different neuronal types to fight against stress, it remains to be determined if the downstream mechanism is also conserved (i.e. up-regulation of Cyp46 or other enzyme involved in cholesterol elimination).

Cholesterol loss is required for survival under high stress

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

Up to this point, we have provided evidences that cholesterol loss triggered by stress increases the activity of TrkRs. However, we did not provide any evidence that this mechanism can be used to improve survival under stress. To test this possibility, we modulated the content of cholesterol in vivo and in vitro and measured cell viability under different experimental conditions. In cultured hippocampal neurons, both the inhibition of cholesterol loss by Cyp46 knock down or cholesterol replenishment increased the apoptosis in old neurons but not in young ones. One interpretation of these findings is that the activation of pro-survival pathways is required only in old cultured neurons that have accumulated high levels of stress. In agreement with the above, lentiviral-mediated knock down of Cyp46 in vivo resulted in neuronal death only in old rats (20 months old) subjected to stress by intraperitoneal injection of kainic acid, to increase glutamate release (Martin et al. 2009). However, neuronal survival was not affected in kainic acid-treated young rats or old rats. A plausible explanation is therefore that cholesterol-triggered survival is required under supra-physiological (or pathological) stressful stimuli.

Cholesterol levels in different regions of the aging brain

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

We reported that maturation and aging are accompanied by a cholesterol decrease in hippocampal membranes. Our results are in agreement with the work of Svennerholm et al. (1991, 1994, 1997) performed in aged human brains and also with data from Perovic et al. (2009) analyzing the levels of expression of cholesterol synthesis and catabolic enzymes in the aged rat. However, others have found that although cholesterol synthesis is decreased in the human hippocampus, the absolute cholesterol content remains stable (Soderberg et al. 1990; Thelen et al. 2006) and even that cholesterol content of the exofacial leaflet of synaptic plasma membranes doubles with age (Igbavboa et al. 1997; Wood et al. 1999). One possibility is that differences arise from the starting material utilized for the measurements. In fact, brain extracts, whether from the hippocampus or any other region, contain a mix of membranes from neurons, astrocytes and oligodendrocytes that can hide the existence of differences. Our measurements in cultured hippocampal neurons provide a reasonable picture of the neuronal situation, as the cultures are largely made out of neuronal cells with very low contamination from astrocytes. In any event, to further assess whether our data reflect neuronal processes occurring in vivo, we have now performed cholesterol measurements in synaptosomes obtained from the hippocampus of 1, 4 and 21 months old mice. Although the presented data correspond to a small number of animals, the results in these synaptic fractions confirm the tendency already observed in cultured hippocampal neurons (Fig. 1).

image

Figure 1.  Synaptic cholesterol levels in mouse hippocampus at different ages. Hippocampal synaptosomes were purified following an established protocol (Nagy and Delgado-Escueta 1984; Dunkley et al. 2008). Then, lipids were extracted and cholesterol and total phospholipids were measured by mass-spectrometry (Bligh and Dyer 1959; Van Veldhoven and Bell 1988; Van Veldhoven et al. 1998). Cholesterol levels were normalized to total phospholipids levels. Hippocampal synaptosomes show an age-dependent decrease in the amount of cholesterol, which was significant between 1 and 21 months. Values correspond to mean ± SD of three animals in each group; *p < 0.05 (Student’s t-test).

Download figure to PowerPoint

All together, it comes clear that changes in cholesterol content do occur during aging, however, not all the brain areas would present changes in the same direction. As a consequence, changes in cholesterol metabolism may influence the survival capacity or performance of certain neuronal populations. In this regard, a direct correlation between cholesterol homeostais and hippocampal activity has been proposed. According to that, impairment of cholesterol synthesis or lipoprotein transport diminishes synaptic plasticity and therefore cognitive functions (Matthies et al. 1997; Blalock et al. 2003; Kotti et al. 2006). Considering the discussed results, previous conclusions on the role of cholesterol in particular physiological or pathological situations (e.g. Alzheimer’s disease) need to be smoothened and re-tested, taking into consideration that cholesterol levels may differ depending on the studied neuronal population.

Conclusion

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References

It is well known that despite accumulation of stress by-products neurons live for the entire lifespan of the individual, indicating that oxidative stress is counterbalanced by strong anti-stress mechanisms. We here show that oxidative stress is a potent inductor of Cyp46 and we also show that this enzyme can help, via cholesterol loss, to improve survival via the TrkB pathway (see Fig. 2). Evidence showing that Cyp46 expression increases in response to ROS accumulation were presented by us (Martin et al. 2008; Iannilli et al. 2009; Sodero et al. 2010) and by others (Ohyama et al. 2006). In further agreement, increases in Cyp46 were found in vivo, in stressful situations (see above). Hence, the first home-take message of our previous work is that stress increases the activity of Cyp46.

image

Figure 2.  Regulation of TrkB activity by Cyp46 up-regulation/cholesterol loss. In young-mature neurons, under low BDNF levels, TrkB receptors are localized in the non-raft fraction of the plasma membrane, in the inactive form. Stressful conditions that induce ROS accumulation (i.e. aging, excitatory neurotransmission) trigger the up-regulation of cholesterol 24-hydroxylase (Cyp46), promoting the elimination of a moderate amount of cholesterol (25–30%) from the plasma membrane. The product of this reaction, 24S-hydroxycholesterol, is a potent repressor of the cholesterol synthesizing enzymes. Under these ‘low cholesterol’ conditions, the composition of the lipid rafts changes, affecting the distribution of membrane proteins. This ‘new rafts’ with low cholesterol, in old-mature neurons, favor the ligand-independent recruitment of TrkB receptors into rafts and its consequent phosphorylation. This, in turn, produces the activation of the PI3K/Akt pro-survival pathway.

Download figure to PowerPoint

A major role of Cyp46 is to make cholesterol soluble and its expression is restricted to the brain (see above). As the brain is a major target of stress and stress increases with age, we did not find surprising that old neurons have low cholesterol levels, both, constitutively (in vivo) or in supra-physiological conditions (in vitro). It would be possible to argue that cholesterol decrease is a typical feature of cultured hippocampal neurons without relevance in situ. However, the same tendency was observed in mouse hippocampal membranes and mouse synaptosomes (Fig. 1 and Martin et al. 2008; Sodero et al. 2010). Then, the use of membranes obtained from hippocampal neurons in vitro seems to be representative of the neuronal situation in vivo. These facts added to the observation that different stressors induced cholesterol loss, make us confident that cholesterol loss in the aged mouse hippocampus reflects the situation of the hippocampus in vivo. As mentioned above, the hippocampal situation might not be the same for other brain structures. This is the second home-take message of our work. To which extent regional variations in cholesterol homeostasis may explain a differential neuronal vulnerability in certain pathologies (like Alzheimer’s disease) needs to be evaluated in the future.

We showed that the Cyp46 pathway is important for neuronal survival in aged animals exposed to stress. Hence, it appears reasonable to hypothesize that any genetic defects in this pathway could lead to pathological aging. This idea is supported by studies reporting single nucleotide polymorphisms in this gene and their association to increased risk of Alzheimer’s disease (Kolsch et al. 2002; Papassotiropoulos et al. 2003; Borroni et al. 2004; see however Desai et al. 2002; Wang and Jia 2007). Considering that TrkB activity is important for survival of differentiated cortical neurons under stress (Gates et al. 2000), it is quite conceivable that intronic mutations could pre-dispose to Alzheimer’s disease through either ‘gain-of-function’ or ‘loss-of-function’ processes. In the first case, by induction of an extra-physiological loss of cholesterol, independently from the presence of stress (see Ledesma et al. 2003; Abad-Rodriguez et al. 2004). In a ‘loss-of function’ scenario, cholesterol loss would not occur upon the appearance of stress, therefore precluding the activation of an important anti-apoptotic pathway like the mediated by TrkB receptors. To dissect how genetic mutations may pre-dispose to disease is another open venue for future research.

References

  1. Top of page
  2. Abstract
  3. Primary hippocampal cultures as model system for certain aspects of brain aging
  4. TrkB activity in the mature brain can occur by BDNF complementary mechanisms
  5. High TrkB activity can be induced by membrane cholesterol reduction
  6. Cyp46 activation determines cholesterol loss in aged neurons in vitro and under acute stress conditions
  7. Oxidative stress determines Cyp46 activation and cholesterol loss in vivo and in vitro
  8. Cholesterol loss is required for survival under high stress
  9. Cholesterol levels in different regions of the aging brain
  10. Conclusion
  11. Acknowledgements
  12. References