Fig. 1. Two-step antioxidant pathway. Superoxide radicals (O-*) generated during oxidative metabolism (e-), are neutralized to water via a two-step process involving superoxide dismutase (SOD) in the first step, and both or either glutathione peroxidase (GPx) and catalase in the second step. An imbalance in this pathway favors the build-up of hydrogen peroxide (H2O2). Fenton-type reactions occur when H2O2 interacts with transition metals such as Fe2+, resulting in the production of noxious hydroxyl radicals (*OH). These radicals initiate rounds of peroxidative damage to molecules such as lipids, via the production of lipid peroxy radicals (LOO*) and lipid hydroperoxides (LOOH). The functional importance of GPxl may reside in its ability to remove both hydrogen and lipid peroxides and neutralize these to water and lipid alcohol (LOH), respectively.

biologically important membranes (45). For these reasons, we investigated the ratio of first to second-step antioxidant enzymes as an indication of the level of antioxidant protection in DS fetal organs. We focused on GPxl expression in DS tissue since this antioxidant enzyme is of greater importance in the brain (46), having wider substrate diversity for the removal of ROS than the peroxisome-specific hydrogen peroxide-reducing antioxidant catalase (47). Furthermore, evidence suggested that some DS tissues up-regulate GPxl in an adaptive response to increased SOD1 levels, while other tissues fail to do so (13,34,48), implying that some DS organs or tissues may be more at risk of peroxidative damage than others.

In the majority of organs studied (namely the brain, heart, lung and thymus), GPxl expression was not significantly different from that of controls, suggesting that no adaptive rise in GPxl occurs in these fetal DS tissues (31,32). In addition, lack of GPxl adaptation affected the SOD1 to GPxl expression ratio in these organs, which was significantly increased (l.5-2-fold) compared with controls. Only the DS liver showed a significant decrease in GPxl expression, however, the ratio of SODl to GPxl in DS fetal livers remained approximately 2: l, since the decrease in GPxl expression was greater than the decrease in SODl expression. Our data therefore showed that all DS fetal tissue tested had an altered SODl to GPxl ratio and suggested that this could initiate oxidative stress-mediated processes in DS (31,32). Based on these data, we hypothesized that oxidant-mediated changes may begin in utero with the potential to continue during the life span of the individual with DS. This may explain the earlier onset of pathologies such as the premature aging and/or neurodegeneration with Alzheimer-like pathology seen in DS individuals, all pathophysiological processes where ROS-mediated damage is implicated.

In this regard, it is noteworthy that DSCRl, a HSA2l gene that we identified in our search for genes associated with the brain phenotypes in DS (26), is redox sensitive and may play a role in neuronal damage. DSCRl expression is induced during cellular adaptation to oxidative stress and is protective against acute oxidative- and calcium-induced stress damage (49). The mechanism involved is unknown but it is assumed that protection occurs via DSCRl's negative regulation of calcineurin and rescue from calcineurin's pro-apoptotic effects (49). However, it has been suggested that long-term depression of calcineurin activity brought about by conditions of oxidative stress and the up-regulation of DSCRl in conditions such as DS and Alzheimer's disease (AD), promotes neuronal damage (50). Mitochondrial dysfunction, which may be a direct consequence of oxida-tive stress, has emerged as a common theme that underlies several neurological disorders including DS (51). Intriguingly, studies in Drosophila have demonstrated that alterations in the abundance of the DSCRl ortholog, nebula, led to mitochondrial dysfunction characterized by decreased ATP, a reduction in the activity of adenine nucleotide translocase (ANT), the ADP/ATP translocator, increased ROS production and a decrease in the amount of mitochondrial DNA (52). These phenotypes were rescued by restoring nebula to mutants. Collectively the data suggest that the redox-sensitive HSA2l-gene DSCRl plays an important role in the regulation of mitochondrial function and integrity. Furthermore, these data imply that the mitochondrial phenotype is a direct consequence of the levels of expression of a single gene and not the consequence of some generalized disturbance as would be predicted by the "developmental instability" hypothesis.


Premature aging

Accelerated aging has been observed in individuals with DS. In particular, rapid or early onset of aging is evident visually as premature graying or hair loss. Detailed biochemical analysis has revealed that DS individuals show a decline in immune responsiveness similar to that seen in older people. Furthermore, alterations in cyclic nucleotide metabolism have been noted in lymphocytes from individuals with DS, which is comparable to lymphocytes of aging humans. Granulovacuolar degeneration of neurons and the appearance of AD pathology, amyloidosis, hypogonadism and degenerative vascular disease have also been noted in DS individuals (53).

In an attempt to understand the genetic and biochemical factors that contribute to premature aging observed in DS, much attention has focused on SOD1, its role within the antioxidant pathway and the consequences of its over-expression in DS. As already described, we demonstrated increased SOD1 expression in most DS fetal tissue investigated, with all organs showing an altered SOD1 to GPx1 ratio (32). In order to understand the effects of an altered antioxidant ratio in DS tissue and whether this contributes to the premature aging of these individuals, it is important to assess the role of antioxidant defense during normal cellular aging. However, the literature is controversial with respect to the activities of the major antioxidants during the aging process, often due to limitations in the methodologies employed. Often no distinction is made between the different iso-forms of the superoxide dismutases, such that SOD1 (the cytosolic and most abundant isoform) and SOD2 (the mitochondrial isoform) are assayed as one. This has resulted in reports of either an increase (54), a decrease (55) or unchanged (46) SOD activity during aging. Likewise, GPx1 activity has been reported to increase (54), or remain relatively unchanged (46,55) in aging brains. Lipid peroxidation is often assayed as a marker of oxidative damage. Again results have been contradictory, with either an increase (56) or a decrease (57) reported in aging brains. Indeed, very few studies have simultaneously investigated the activities of SOD1, GPx1, catalase and lipid peroxidation, which becomes important if discrepancies such as differences in sex, species, strain and age are to be eliminated.

In order to address all of these parameters in the same tissue, we examined the levels of SOD1, GPx1 and catalase, and the extent of lipid peroxidation during aging in murine brains (58,59). We showed that SOD1 activity was significantly increased during the aging process, while the activities of GPx1 and catalase remained largely unchanged. This contrasted with other murine organs where increased SOD1 activity was accompanied by an increase in either or both GPx1 and catalase. Our data therefore clearly showed that most organs compensate for the elevated SOD1 levels during aging, while the brain failed to do so. Importantly, lipid peroxidation was also highest in the brain and this correlated with the lack of compensation in this organ, while those organs that adapted to the increased SOD1 levels by up-regulating GPx1 and/or catalase showed reduced peroxidative damage (58,59).

Our data therefore support the notion that an organ such as the brain is most vulnerable to oxidative insult that may arise as a consequence of an altered antioxidant ratio, while tissues with an unaltered ratio experience less peroxidative damage. Based on our study where all DS fetal organs investigated have an altered antioxidant ratio (32), and the fact that an altered ratio is associated with increased lipid damage, we hypothesized that all DS fetal organs may be vulnerable to peroxidative attack. Furthermore, it is not known why certain organs such as the murine brain failed to adapt to the increased SOD1 levels while other murine organs adapted by up-regulating GPx1. Interestingly, the adaptive response in DS has been noted in certain adult DS tissues (34,48), but not in fetal brains (13,32). It is possible that adaptation only occurs in adult DS tissue and only those organs able to respond to the higher SOD1 levels. Investigations of this type are clearly limited to available tissue from DS individuals such as fibroblasts and blood cells, where adaptation by GPx1 has indeed been observed. However, it is conceivable that an organ such as the brain lacks the ability to adapt (as was seen during murine aging (58,59)), making the DS brain highly susceptible to peroxidative attack that begins in utero and continues throughout the lifetime of the DS individual.

Subsequent studies in our laboratory have shown that failure to adapt to increased SOD1 levels (by not up-regulating GPx1) has profound aging-like effects on cells in culture (60). Our data showed that SOD1-transfected cells that had an elevation in the SOD1 to GPx1 ratio produced higher levels of hydrogen peroxide and exhibited well-characterized markers of cellular senescence, e.g. slower cellular proliferation and altered cellular morphology. These changes were similar to those seen in cultured fibroblasts obtained from DS individuals. Furthermore, treatment of normal cells with hydrogen peroxide was able to mimic these effects in culture, suggesting that hydrogen peroxide mediated these senescent-like changes (60). We also identified that Cip1, a known senescent gene, is up-regulated by hydrogen peroxide and increased in DS cells, therefore establishing that senescent-like changes in DS are mediated via ROS-specific pathways that involve known senescence genes. Analysis of cell lines derived from mice deficient in GPx1 further confirmed that an altered SOD1 to GPx1 ratio resulted in senescent-like changes and that these effects occurred as a result of ROS-mediated events (61). Collectively, our data strongly suggest that an altered SOD1 to GPx1 ratio in DS leads to senescent-like changes and that this is mediated in part by hydrogen peroxide.


Compared with other organs, the brain is most vulnerable to ROS-induced damage. A number of contributing factors exist, including: (i) increased ROS formation as a consequence of the high rate of oxygen consumption by the brain (62); (ii) high iron levels in some brain regions that catalyze ROS formation via Fenton-type reactions (63) (see Fig. 1) and (iii) the high prevalence of polyunsaturated fatty acids (PUFAs), particularly in cerebral neuronal membranes, which are known targets for peroxidative events (64).

Therefore, protection against ROS-mediated damage in the brain is paramount for the maintenance of cellular integrity and function.

An altered SOD1 to GPx1 ratio in the DS brain may affect the survival and function of both neuronal and non-neuronal cells of the brain. In vivo evidence for an impact on the DS brain as a whole comes from Brooksbank and Balazs (13) who demonstrated that the altered antioxidant ratio in DS fetal brains is accompanied by increased lipid peroxidative damage. Strong in vitro evidence for an effect on neuronal cells comes from Busciglio and Yankner (4), who showed that the increased neurodegeneration of DS-cultured cortical neurons was accompanied by increased lipid peroxidation and apoptosis. In addition, these effects were mediated via hydrogen peroxide since addition of compounds such as N-acetylcysteine and catalase (but not SOD1) prevented the degeneration of these cultured DS neurons. Furthermore, studies in mice transgenic for SOD1 have provided invaluable information on the role of an altered antioxidant ratio in various pathologies associated with DS. Over-expression of SOD1 enhanced ischemic reperfusion injury in fetal brains, suggesting a deleterious role for elevated ROS in brain development (65). SOD1 transgenic mice also developed morphological and biochemical changes at tongue (66,67) and hindlimb (68) neuromuscular junctions, which are similar to those seen in individuals with DS. Furthermore, SOD1 over-expression led to a chronic pro-oxidant state in the brain, as evident by increased levels of oxidized glutathione and altered calcium home-ostasis. SOD1 over-expressing neurons were also more susceptible to kainic acid-mediated apoptotic cell death (69). Constitutive elevation of SOD-1 activity also exerted a major effect on neuronal excitability, which in turn, affected the H2O2-mediated hippocampal ability to express long-term potentiation (LTP) (70). These studies led to the proposal that elevated SOD1 causes an increase in H2O2 which diminished LTP and cognitive deficits in these mice (71). Finally, over-expression of SOD1 in rat PC12 cells showed impaired neurotransmitter uptake, resulting in diminished transport of biogenic amines into chromaffin granules (72). Since neurotransmitter uptake plays an important role in many processes of the central nervous system, SOD1 gene-dosage with its inevitable alterations in antioxidant balance, may contribute to the neurobiological abnormalities of DS.

Data from our laboratory, investigating mice deficient in GPx1, provide further evidence that an altered SOD1 to GPx1 ratio affects neuronal integrity and function. First, we demonstrated that GPx1-deficient neurons were more susceptible to H2O2-mediated toxicity (73) and that cell death occurred via apoptotic processes (74). Second, we showed that mice deficient in GPx1 were more susceptible to neuronal apoptosis following mid-cerebral artery occlusion (a stroke model of ischemia/reperfusion) (75) leading to a greater cerebral infarction in these mice. This was accompanied by accelerated caspase-3 activation, a clear indication that apoptotic pathways are affected by an altered antioxidant ratio. Similarly, we showed that GPx1-deficient mice were more susceptible to neuronal apoptosis in a murine model of head trauma (76). In this instance, an initial neuro-inflammatory response was accelerated in GPx1-deficient brains.

In an attempt to understand how the lack of GPx1 enhances susceptibility to apoptosis, our group investigated the functionality of the phospho-inositide 3-kinase [PI(3)K]-Akt pathway in our model of cerebral ischemia-reperfusion injury as well as in primary cultured neurons from GPxl-deficient mice. We chose to study the PI(3)K-Akt pathway since it is known to be an important signaling cascade critical for the protection against neuronal cell death (77,78) and it is also known to be affected by growth factors and H2O2 (77). Importantly, our study showed that aberrant Akt phosphorylation occurred in GPxl-deficient neurons following treatment with nerve growth factor (NGF) or H2O2 (74). In addition, levels of the upstream PI(3)K subunit p85 were also reduced in GPxl-deficient neurons. This resulted in reduced Bad phosphorylation/activation, highlighting that the downstream functionality of the PI(3)K-Akt pathway is disrupted in GPxl-deficient neurons. Under normal conditions, activated Bad is sequestered in the cytoplasm, preventing its interaction with pro-apoptotic Bcl-XL. In this manner, mitochondrial cytochrome c is not released, downstream caspases are not activated and apoptotic processes are averted. Lack of Bad activation, on the other hand, as seen in GPxl-deficient neurons (and therefore by inference, increased availability of Bcl-XL) is suggestive of a lack of protection by this pathway in GPxl-deficient neurons. Our studies clearly show that an imbalance in the antioxidant ratio affects the ability of signal transduction pathways to coordinate pro-survival responses, with direct implications for the compromised survival of neurons in the heightened oxidative state of DS. It is therefore reasonable to deduce from our studies that perturbations in key anti-apoptotic mechanisms, as a consequence of an altered antioxidant ratio, may have significant implications in neuropathologies such as DS.

An altered antioxidant balance, leading to increased oxidative stress, may also affect the function of non-neuronal cells of the brain. In particular, astrocytes play an important role in the glutamate-glutamine cycle, the dysregulation of which has been implicated in a number of neurological disorders (79). The astrocyte-specific enzyme, glutamine synthetase (GS; EC 6.3.l.2) is responsible for the replenishment of L-glutamine, the precursor required by neuronal cells for the regeneration of the excitatory neurotransmitter L-glutamate. In a recent study, we investigated the impact of oxidative stress on the function of glutamate synthetase in astrocytes derived from GPxl-/- mice (80). Lack of GPxl in the presence of a chronic oxidative stress led to reduced GS function, implying that GPxl contributes to the protection of this important neurotransmitter-regenerating enzyme. Clearly, increased oxidative stress in DS brains may impact upon the function of non-neuronal cells such as astrocytes, and this may be an additional mechanism responsible for the impaired brain function in DS.

Finally, a further study by our group worthy of note involves the HSA2l-specific transcription factor ETS2. In initial studies we showed that ETS2 expression, which is increased 5-7-fold in DS fibroblasts, is induced by hydrogen peroxide (81). Since it is known that hydrogen peroxide mediates the increased rate of apoptosis of DS cells (4), our data implicated ETS2 in the regulation of oxidant-mediated apoptosis. Furthermore, we showed that moderate over-expression of ETS2 increased apoptosis of primary neuronal cultures derived from ETS2 transgenic mice via a mechanism that involved the activation of caspase-3 and was dependent on the tumor-suppressor gene p53 (82,83). These data implicate ETS2 in the regulation of oxidant-induced apoptosis/ neurodegeneration and provide a possible rationale for both the greater than gene-dosage increase in ETS2 protein level in DS tissues, and the elevated rate of apoptosis in DS cells.

Alzheimer-like pathology in Down syndrome

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by progressive memory loss, intellectual function and cognitive abilities (2). In the general population, AD mostly occurs during the sixth decade of life and approximately 10% of individuals aged 65 and above show cognitive signs of AD (84). In DS, however, all patients develop Alzheimer-type neuronal pathology by the third to fourth decade of life (2). Plaques and tangles that develop within the brain of DS individuals are virtually identical to those seen in patients with AD (3), consisting of bundles of uniform proteins that appear as paired helical filaments on electron microscopic examination (85). Aß peptide, which is a normal product of cell metabolism derived from the ß-amyloid precursor protein (APP), is overproduced in both the brains of individuals with AD and DS, although the mechanisms leading to the overproduction of Aß in these pathologies apparently differ. APP is a HSA21 gene, and over-expression in DS at levels greater than expected by gene dosage (86), is related to both trisomy of the gene and modulation by other factors such as oxidative load. Over-expression in AD is attributed to aberrant APP gene processing (87) with the majority of gene mutations in familial early-onset AD patients occurring in presenilin genes that mediate y-secretase (protease) APP cleavage leading to increased Aß formation (88). Irrespective of the mode of increased Aß production, once aggregated this insoluble protein forms part of the fibrillar neuritic plaque that is thought to promote neuronal degeneration (89). In persons with DS, soluble Aß peptides appear in the brain decades before the extracellular deposition of neuritic plaques. These soluble amyloido-genic peptides accumulate intraneuronally and are secreted extracellularly. Indeed, their appearance has been reported in the brains of fetuses with DS (90).

There is now strong evidence that both amyloid ß-peptide and oxidative stress play an integral role in the neurotoxicity of plaques in AD and DS individuals (89,91). In vitro studies have shown that Aß peptides are toxic to a wide variety of neuronal cell-types by increasing mitochondrial ROS production (92-94). This results in disruption of Ca2+ homeostasis (89,95) and renders neurons more susceptible to excitotoxicity and apoptosis (96,97). In addition, Aß peptides bind to RAGE, the receptor for advanced glycosylation end products, effecting an oxidant-sensitive nuclear factor kappaB (NF-kB)-dependent inflammatory response, which includes up-regulation of interleukin-1 (IL-1), interleukin-6 (IL-6) and tumor necrosis factor-a (TNF-a) (98). Furthermore, oxidative stress per se, exacerbates Aß aggregation (99) and in doing so, elicits a pathological oxidative stress/Aß peptide cycle, ultimately causing neuronal damage. Several studies have implicated hydrogen peroxide as the major source of oxidative stress generated by Aß peptides in in vitro systems (94,100). It is therefore most likely that H2O2 is involved in the deposition of Aß, and is responsible for mediating the oxidative stress/Aß cycle. Evidence for this comes from studies where cells with intrinsically higher levels of peroxide-removing antioxidant enzymes displayed increased resistance to Aß toxicity (101). Additional evidence comes from a recent study by our group where GPx1-deficient neurons were more susceptible to Aß toxicity than wild-type counterparts, implying that an altered antioxidant balance and H2O2 play a role in neuronal Aß-mediated toxicity (102) Furthermore, we showed that treatment of GPx1-deficient neuronal cultures with the free radical scavengers N-acetylcysteine (NAC) and the GPx mimic ebselen, abrogated the cytotoxic effect of A3. Our findings also nicely complement recent data of Barkats et al. (103), who show that mice over-expressing GPxl display greater resistance to A3-mediated toxicity. In addition, oxidative stress and hydrogen peroxide in particular, has been shown to affect the clearance of A3 peptides, since insulysin and neprilysin, two proteins required for the degradation of A3 are inactivated by oxidative mechanisms (104). Finally, evidence that other factors are required in addition to increased APP levels for A3 deposition in DS and AD, comes from studies where elevated levels of APP alone were insufficient to produce amyloid deposition. Stably transfected cells that over-expressed APP exhibited increased levels of A3 but did not display any extracellular A3 deposits (105), and mice over-expressing APP did not exhibit extracellular deposits of fibrillar A3 and/or neuronal degeneration (89). In the light of this data, it is interesting to note that the APP promoter region contains a heat shock element (HSE) that is redox sensitive (106), thereby linking oxidative-regulatory mechanisms and APP production. Data from our group also implicate the redox-sensitive transcription factor ETS2 in the regulation of the APP gene via specific ETS-binding sites in the beta-APP promoter (107).

It is tempting to speculate that the increased oxidative stress, which contributes to the Alzheimer-like pathology in DS, arises as a consequence of the altered antioxidant ratio. Compelling evidence for this comes from a study investigating the consequences of a combined increase in APP and SOD1 in a double-transgenic (tg)-APP-SODl mouse (108). In addition to having severe impairment in learning and long-term memory, the brains of aged (tg)-APP-SODl mice demonstrated an accumulation of membrane-bound high-molecular-weight APP species, severe morphological damage including lipofuscin accumulation and mitochondrial abnormalities that were far greater than changes seen in APP or SOD1 mono-transgenic mice. Thus, the combined elevation of two HSA21 genes in tg-APP-SODl mice (one affecting the antioxidant ratio and the other the production of APP) was required to elicit the greatest effect on age-dependent alterations in morphological and behavioral functions. In addition, as reported earlier, our data showed that an altered SOD1 to GPxl ratio led to increased susceptibility of GPxl-deficient neurons to H2O2-mediated toxicity (73) and that this was associated with increased lipid damage (58,60), implying that an altered antioxidant ratio has neuropathological consequences.

Based on the available evidence, it is attractive to propose the following scenario in which an altered SOD1 to GPxl ratio, together with aberrant APP processing leads to increased A3 deposition in DS brains: oxidative stress, due to the elevated SOD1: GPx ratio, leads to higher levels of APP via the redox-sensitive HSE and/or ETS2-mediated transcriptional induction of the APP gene. Indeed, this might explain the greater than gene dosage increase in APP (approximately 4-5-fold increase) seen in DS individuals. In addition, as discussed previously, ETS2 is also redox sensitive (81), so an altered antioxidant ratio could potentially up-regulate ETS2, which in turn mediates increased APP expression. The increased APP production, in turn leads to increased formation of A3 (due to aberrant APP processing) and fibrillar A3 amyloid deposits, which results in further oxidative stress, disruption of calcium homeostasis, mitochondrial dysfunction and consequently increased apoptosis and neurodegeneration (Fig. 2).


(senescence/apoptosis/ necrosis)

■ftROS formation

Down syndrome

■premature aging ■neuronal cell death ■Alzheimer-type pathology ■ mental retardation: defective neurotransmitter uptake into neurons

-ftTNFa, IL-1 Inflammation IL-6

-ftTNFa, IL-1 Inflammation IL-6

Fig. 2. The consequences of an altered SOD1 to GPx1 ratio in Down syndrome (DS). An imbalance in the SOD1 to GPx1 ratio results in the build-up of reactive oxygen species (ROS). ROS have been implicated in numerous pathologies that also occur as part of the DS phenotype, such as premature aging, neurological disorders and Alzheimer's disease (AD). An understanding of how ROS contribute to individual pathologies aids in the understanding of more complex situations such as DS. In particular, ROS have been implicated in the formation of Ap, the neurotoxic protein found in AD plaques, via a mechanism that includes the redox-sensitive transcription factor ETS-2 and the P-amyloid precursor protein (APP). This process, in turn, generates more ROS, thereby further fueling the build-up of A|P Increased ROS also affects inflammatory pathways via activation of the redox-sensitive transcription factor NF-kB, which in turn increases expression of the interleukins, IL-1, IL-6 and tumor necrosis factor, TNF-a. This in turn generates more ROS, thereby enhancing inflammatory processes.

Neurological disorders

4 ETS2^=> ft APP expression; ■ft APP expression via redox-sensitive HSE; ■aberrant APP processing i-> ft Ap

ROS formation ^ ■Pyramidal neuronal cell death

The Latest Anti Aging Treatments

The Latest Anti Aging Treatments

Are You Striving To Look And Feel Youthful? Wish You Could Add 20 Years To Your Life? Discover the Secrets to a Longer, Healthier Life With This Fantastic Anti-Aging Resource. You might be feeling and looking great now, but have you ever thought about what youll feel and look like several years from now? Have you ever considered that the choices you make today directly influence how well you age?

Get My Free Ebook

Post a comment