Pii: s1471-4914(01)02017-2

TRENDS in Molecular Medicine Vol.7 No.8 August 2001 unlikely that neuronal loss in neurodegenerative How do neurons die
diseases is solely accomplished by apoptosis. Anyproposed mechanisms of neuronal death shouldexplain this extraordinarily slow time course.
in neurodegenerative Morphological and molecular hallmarks of individual
neurodegenerative diseases
diseases?
Each disease has its own hallmarks. These hallmarksare the obvious choice as parameters for the study ofspecific neural death.
Ichiro Kanazawa
Senile plaques and neurofibrillary tangles in corticalneurons in AD Given that neurons are post-mitotic cells, their life span is generally long
Extracellular senile plaques (SP) and intra- enough to reach that of humans. However, sometimes neurons die without
neuronal neurofibrillary tangles (NFT) are cardinal recognizable causes, as a result of a process called neurodegeneration. Apart
pathological hallmarks of AD. The main chemical from when gene mutations can be correlated with disease, it is difficult to
component of the core of SP is amyloid β protein pinpoint molecules that are responsible for neuronal death. Therefore,
[Aβ, a mixture of Aβ40 and Aβ42 proteins, which are neurons living in a ‘sick state’ for many years might reveal important
produced by cleavage of an amyloid β protein information about neuronal death. Systematic and extensive single-neuron
precursor (APP) by secretases]5. Presenilins, which analysis of ‘sick’ neurons is expected to provide clues to the mechanisms of
have recently been identified as γ-secretases, are neurodegeneration. Moreover, the elimination of putative triggering and
mutated in a subset of early-onset familial AD promoting factors involved in neurodegenerative disease might prevent
(Refs 6,7). Given that Aβ added to cultured neurons disease progression.
is toxic to the cells8, neuronal death is expected tooccur by intracellular accumulation of Aβ. The Neurodegenerative disorders are characterized second hallmark of AD is the intracellularly clinically by insidious onset and slowly progressive deposited NFT, which is a polymerized, argyrophilic course, and are frequently hereditary. Pathologically, abnormal structure composed of hyperphosphorylated these diseases share a common feature: the selective tau protein9. By analogy with studies of Down’s loss of a particular subset of neurons for unknown syndrome, AD pathology might begin with the reasons [e.g. cerebral cortical neurons in ALZHEIMER’S formation of SP and years later proceeds to the DISEASE (AD) (see Glossary), substantia nigra neurons formation of NFT (Ref. 10). In this respect, it is in PARKINSON’S DISEASE (PD), spinal motoneurons in worth noting that the tau-phosphorylating enzyme AMYOTROPHIC LATERAL SCLEROSIS (ALS), and striatal (tau phosphokinase, TPKI) could be one of the small neurons in HUNTINGTON’S DISEASE (HD)].
missing links between SP and NFT. Indeed, Apoptosis has recently been implicated as a possible the activation of intracellular TPKI is induced by the mechanism for neuronal death in neurodegenerative extracellular Aβ (Ref. 11). Neuronal loss in the diseases. However, there is no direct and convincing superior temporal gyrus of AD patients exceeds evidence of apoptosis in human brains, and the the number of NFT-positive neurons by more than mechanisms of neuronal death in neurodegenerative sevenfold1. Therefore,the majority of neurons can diseases are still unknown. Here, I will discuss die without developing NFT. Thus, although Aβ in several proposed mechanisms of neuronal death in individual diseases. In addition, I will put forward molecules for the understanding of the AD the concept of a long-standing ‘sick state’ of pathogenesis, there is still a gulf between hallmarks remaining neurons and the possible underlying mechanisms of neuronal ‘sickness’.
Neuronal loss in neurodegenerative diseases is an
The cardinal pathological hallmark of PD is the extremely slow process
appearance of hyaline-like intracytoplasmic Each neurodegenerative disease has its own clinical inclusions, Lewy bodies (LB). LB are found in the course. AD, PD and HD, for example, begin remaining dopaminergic neurons in the substantia gradually and progress slowly for more than nigra and other nuclei. Following the identification of 10–20 years. By contrast, ALS progresses rapidly an α-synuclein gene mutation as the cause of and the disease process usually lasts only 2–3 years.
dominantly inherited rare PD (Ref. 12), α-synuclein Corresponding to the clinical course, the time course has also been established as the major component of Ichiro Kanazawa
of neuronal loss is slow in AD and PD, and relatively LB in sporadic PD. Indeed, a mutation in the gene rapid in ALS (Refs 1–3) (Fig. 1). The speed of encoding α-synuclein leads to the loss of neuronal loss in neurodegenerative diseases is much dopaminergic neurons and intra-neuronal inclusions slower than that of apoptotic neuronal loss in in a Drosophila model of PD (Ref. 13). Normal developing nervous systems4. It is, therefore, α-synuclein is localized in the presynaptic terminals, http://tmm.trends.com 1471-4914/01/$ – see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S1471-4914(01)02017-2 TRENDS in Molecular Medicine Vol.7 No.8 August 2001 Glossary
Alzheimer’s disease (AD): Neurodegenerative disease that begins
with insidious memory loss in late life. Subsequently, language and visiospacial impairments occur. Frequency in over 65 years is approximately 3–5%. AD is predominantly sporadic, with <20%being autosomal-dominantly inherited. Amyotrophic lateral sclerosis (ALS): Characterized by muscle
atrophy or weakness in hands, feet or tongue from age 20 onwards, progressing rapidly. Respiratory impairments can be life threatening. Frequency is about 1–3 cases per 100 000.
CAG repeat: Glutamine-encoding tandemly repeated codon found
in the human genome. Abnormal expansion of CAG repeats (to >40) can result in disease-causing polyglutamine stretches.
Huntington’s disease (HD): Initial symptoms include chorteic
involuntary movement or intellectual deterioration in the 30s or 40s, but the disease typically progresses slowly. The HD-causing mutation is an expansion of a CAG repeat in the IT15 gene.
Parkinson’s disease (PD): Neurodegenerative disease that begins
with insidious tremor at rest, or slowness in movement on one side of the body in middle to old age. Frequency is about 100 cases per 100 000 population. More than 90% of the patients are RNA editing: A mechanism of post-transcriptional modification. A
single nucleotide can be substituted in the mRNA sequence; for example, CAG (Gln codon) is substituted to CGG (Arg codon).
Fig. 1. Superimposed
binds to synaptic vesicles, and is transported by the axonal flows. The mutation in the α-synuclein gene renders the protein devoid of vesicle-binding activity and promotes accumulation-forming β-sheet transgenic mice containing the mutated SOD1 structure14. However, the precise role of this protein gene18. Moreover, given that the familial type is a in neurodegeneration is still unclear. Recently, the rare subset of ALS, SOD might not have a role in parkin gene was identified as being responsible for an autosomal-recessive form of juvenile-onset PD.
Parkin is thought to be a substrate for ubiquitination Intranuclear inclusions in striatal small neurons of HD leading to protein degradation by the proteasomal HD is caused by an expansion of CAG REPEATS located complex15. Because LB are not formed in this in the coding region of the IT15 gene, whose product recessive PD syndrome, a possible correlation is named huntingtin. Abnormal huntingtin, between parkin protein and the pathogenesis of therefore, has an unusually long glutamine tract19.
sporadic PD is uncertain. Furthermore, the normal Intranuclear inclusion bodies were found in the function of parkin is also yet to be clarified.
striatal neurons of transgenic mice expressing the CAG repeat containing exon 1 of the huntingtin Intracellular inclusions in spinal motoneurons in ALS gene20. The inclusion bodies are also found in the The Bunina body in spinal motoneurons is the most HD striatum and cortex. Inclusion bodies are well known intracellular inclusion body associated aggregates of a truncated form of protein containing with ALS, but is still not fully characterized. Other a polyglutamine stretch, ubiquitin, glyceraldehyde- inclusion bodies in motoneurons of ALS are, unlike 3-phosphate dehydrogenase (GAPDH) and many LB in PD, not uniform and are described by various other proteins. However, there does not seem to be a names such as argyrophilic, hyaline, conglomerate correlation between the formation of inclusion and skein-like inclusions. Most of them are, bodies and neuronal death in cultured neurons that however, an accumulation of phosphorylated express abnormal huntingtin21. By analogy with neurofilaments16. These findings, and the presence of other polyglutamine diseases, the incorporation of axonal spheroids suggest that ALS might be strongly related to the disturbance of neurofilament stretch into the nucleus itself, rather than the function. Recently, mutations in the intracellular formation of aggregates, is now thought to cause Cu2+–Zn2+-dependent superoxide dismutase (SOD1) neuronal death through disturbing normal gene were discovered in rare familial ALS (Ref. 17).
functions of transcription factors22. The actual Spinal motoneurons from familial ALS patients relationship between the disturbed gene expressions frequently bear intracellular inclusions that are and the consequent neuronal death is still a immunoreactive for an antibody against SOD1.
matter of debate. In summary, hallmarks of Because SOD1 acts as a detoxifier of free radicals, a neurodegenerative diseases are valuable clues for mechanism of neuronal death is expected to be understanding the pathogenesis of the disease.
related to the loss of SOD activity. However, there is However, it is important to recognize that disease no positive correlation between the gene defect and hallmarks are not necessarily parameters for motoneuron loss, not only in patients but also in TRENDS in Molecular Medicine Vol.7 No.8 August 2001 Molecular mechanisms of ‘sick state’ of neurons
Whatever the initial trigger of neuronal death inneurodegnerative diseases is, consequent events canproceed insidiously, gradually or episodically. Eachneuron has its own optimal intraneuronalbiochemical conditions such as intracellular pH,water content, concentrations of oxygen, glucose,ATP, second messengers and Ca2+ ions. In ‘sick’neurons, these conditions might deviate slightlyfrom the optimum, without exceeding the life-threatening limit for neurons. It is possible,therefore, that a long-standing unfavorable living condition might make neurons consume their energyinsidiously and after many years lead to theneuronal death. Possible mechanisms of ‘sickness’ ofneurons are summarized in Fig. 3.
Aging processBecause aging proceeds insidiously from middle life (for a review see Ref. 29), similar toneurodegenerative diseases, one might hypothesizethat neurodegenerative diseases are caused by anaccelerated aging process. Indeed, a certain number of neurons are lost with age30: nigral neurons are Fig. 2. Morphological
‘Sick neurons’
most severely affected, cortical neurons next and In PD patients, less than 20% of nigral neurons spinal motoneurons least. However, the speed of remain 20 years after onset of the disease. Because neuronal loss in AD and PD is notably faster than they are destined to die, these remaining neurons might provide important insight into neuronal degeneration. Some of the remaining neurons show ‘degenerative changes’ in terms of size, shape and significantly increased in the brains of elderly morphology of neuronal soma and dendrites.
people. Nonenal might contribute to membrane Although these changes are not specific, they might damage and increased susceptibility to free radicals represent definite signs of ‘sickness’ of neurons. and consequently lead to serious disturbance of For example, the density of dendritic branches of neuronal structures and functions31, and 8-OHdG most cortical neurons becomes coarse even in the might be a marker of DNA oxidation. Moreover, early stage of AD (Ref. 23). In addition, significantly inactive enzymes, oxidized proteins and structurally reduced numbers of dendritic spines and synaptic altered proteins increase with age in the brain29. The terminals were noticed in AD cortex24. These findings amino acid groups of proteins non-enzymatically support the decreased synaptic function in AD brain.
react with glucose or other monosaccharides, a form In the substantia nigra of PD patients, the remaining of post-translational modification. The products of neurons show condensation of cytoplasm and nuclear this reaction further produce, through oxidations or indentation of neurons25. Quantitatively, 4–40% of dehydrations, more complex protease-resistant dopaminergic nigral neurons in PD were reported to large molecules, for example, advanced glycation show apoptotic cell fragmentation and autophagic end-products (AGE), which increases with age. AGE degeneration26. Indeed, our laboratory observed that promotes inter- and intra-molecular crosslinking, nearly 50% of remaining nigral neurons exhibit and disturbs normal protein function32. It is cellular shrinkage, cytoplasmic and nuclear possible, however, that the aging process itself is not condensation, and nuclear deformities (Fig. 2). sufficient to kill neurons, but acts as a maintaining In ALS, remaining spinal motoneurons shrink to factor of the ‘sick state’ of neurons.
~70–80% in size27. By contrast, the remaining striatalneurons in dominantly inherited HD show little shrinkage, but frequently show nuclear indentation28.
Oxidative stress (reviewed in Refs 33,34) is caused by When taking the slow time course of neuron loss into the enhanced production of harmful cellular oxidants: consideration, surprisingly large numbers of neurons free radicals (e.g. hydroxyl radical (.OH), superoxide
survive for more than 3–10 years in their seemingly (O –), hydrogen peroxide (H O ), nitrogen oxide (NO) atrophic and/or deformed state. These particular and peroxynitrite (ONOO–)], or a failure of protective ‘sick’ neurons might be maintained at a lower level mechanisms, including superoxide dismutase (SOD) or than normal in terms of metabolism and function glutathione peroxidase. Free radicals can enhance membrane permeability to various molecules through TRENDS in Molecular Medicine Vol.7 No.8 August 2001 Fig. 3. A hypothetical
peroxidation of membrane lipid, and lower the level of neuronal activity. Of course, there are two protective Glutamate is the most abundant excitatory systems against free radicals in living cells, neurotransmitter in the brain, and almost every (1) enzymes converting radicals into harmless neuron expresses glutamate receptors, either compounds (e.g. SOD and glutathione peroxidase), and permeating ions directly (AMPA/KA and/or NMDA (2) non-enzymatic antioxidants (e.g. ascorbic acid or types) or indirectly (metabotrophic type). An increase tocopherol). If processes for free radical production are of extracellular glutamate produces prolonged somehow enhanced and protective processes reduced, depolarization of neurons, inducing prolonged Ca2+ neurons could die. Indeed, there is evidence that free influx into glutamate-receptive neurons, which then radicals play a role in neuronal death not only in leads to neuronal death (i.e. excitotoxicity )37,38. The ischemic brain lesion but also of AD, PD, ALS and HD.
role for excitotoxicity in neuronal degeneration has In PD, free radicals are easily produced with the help of been extensively studied in ALS and HD. Although Fe2+ in the course of the metabolism of dopamine.
the concentrations of glutamate in the spinal cord Therefore, dopaminergic neurons are always exposed and the brain of sporadic ALS patients are not to free radicals. Evidence of a role of oxidative stress in increased39, the predominant high-affinity AD and HD is also accumulating35,36.
TRENDS in Molecular Medicine Vol.7 No.8 August 2001 specifically in astrocytes is lost in the ALS spinalcord. This might be a result of aberrant mRNA caused by splicing errors40, causing an increase ofavailable glutamate in the peri-motoneuronal environment. mRNA for the glutamate receptor 2(GluR2) subunit, which strongly regulates Ca2+conductance of the AMPA/KA receptor, is editednormally at the Gln/Arg residue in the subunitassembly. Recently, in the ventral horn of ALS patients, GluR2 RNA EDITING was shown to besignificantly reduced41. This can lead to a continuousCa2+ influx through AMPA/KA receptors, therebymaking motoneurons vulnerable to various endogenous or exogenous adverse insults. Indeed, ina mouse model of cerebellar ataxia, a causativemutation in the gene (GluRδ2) in the lurcher mouseleads to continuous Ca2+ influx and cerebellar Reduced protein synthesisUsing DNA microarray techniques, a recent study with a mouse model of HD revealed that in the earlystage of the disease, the brain expresses reducedlevels of mRNA of certain receptors and second messengers, but not of mitochondrial proteins orapoptosis-related proteins43. Indeed, the dopamine- Fig. 4. Ultramicro RT-PCR analysis of genes expressed in a single
synthesizing enzyme tyrosine hydroxylase (TH) and human neuron. Brains were obtained at autopsy. Frozen 20µm sliceswere cut by cryostat-microtome. After a freeze-dry procedure, a single its mRNA has been reported to be reduced in the neuron was dissected from the substantia nigra using an excimer laser remaining nigral neurons of PD (Ref. 44).
microdissector. PCR primers were designed to amplify the tyrosine Preliminary single neuron analysis also showed that hydroxylase (lane 1), dopa-decarboxylase (lane 2), α-synuclein (lane 3) the remaining nigral ‘sick’ neurons in PD patients and the ubiquitously expressed GAPDH (lane 4) genes. RT-PCR products appeared as bands in a healthy control (C) and a Parkinson’s definitely express GAPDH mRNA at normal levels disease patient (PD, three different neurons; a–c). N indicates a but express less than normal levels of TH, dopa negative control without RNA templates (Jeong, S.M. et al., decarboxylase and α-synuclein mRNAs (Fig. 4).
unpublished) All four proteins examined are expressed in a nigral These lines of evidence suggest that it is important to ‘sick’ neuron of PD patient (a), whereas dopa-decarboxylase is extremely reduced in another (b). The third ‘sick’ neuron (c) only know the overall expression profile of neurons in ‘sick state’ to clarify the mechanism of ‘sickness’.
dysfunction in AD could be successfully reversed by Therapeutic implications of ‘sick’ neurons
this treatment. Apart from the possible reversal of Although the ‘sick state’ of neurons is generally ‘sickness’ of neurons, there are several lines of regarded as irreversible, one could speculate that therapeutic trials, either experimental or clinical, for ‘sickness’ of neurons could go back to the normal state, neurodegenerative diseases. First, the fetal nigral if triggering or promoting factors for neuronal tissues were transplanted to the striatum of PD damage were eliminated. This assumption is based on patients and a part of dopaminergic function was a recent report of HD transgenic mice model using a improved. Second, a glial cell line-derived tet/off system45. The system makes it possible to turn neurotrophic factor (GDNF) introduced directly into off the expression of a transgene with oral substantia nigra or indirectly by vectors provided administration of tetracycline analogs at any age protection of nigral neuronal death and functional after birth. The damage of the striatal neurons of this recovery of the nigra in the experimental model of PD particular mouse model is definitely reversed by (Ref. 48). Third, a recent experimental study inhibition of the continuous expression of the mutant demonstrated that grafts derived from human fetal HD-causing gene. If extrapolated to HD and other striatal tissue can survive, develop, and are polyglutamine diseases in humans, the ‘sick’ neurons could go back to normal by inhibiting the expression transplantation into a patient with HD (Ref. 49).
of the mutant gene. In addition, vaccinations with Aβ Finally, using mesenchymal cell-derived factor(s), peptide were reported to reduce amyloid deposition in mammalian ES cells successfully differentiated into a transgenic mouse model of AD (Ref. 46). Moreover, neurons, particularly dopaminergic neurons50. This the learning ability of Aβ-vaccinated mice was found achievement opens the door towards a cure of PD, and to be superior to that of non-immunized mice47.
hopefully this therapeutic principle can be applied to Therefore, it is possible to speculate that the cognitive the other neurodegenerative diseases.
TRENDS in Molecular Medicine Vol.7 No.8 August 2001 References
18 Gurney, M.E. et al. (1994) Motor neuron 34 Cohen, G. and Werner,P. (1994) Free radicals, 1 Gomez-Isla, T. et al. (1997) Neuronal loss degeneration in mice that express a human Cu,Zn oxidative stress, and neurodegeneration. In correlates with but exceeds neurofibrillary superoxide dismutase mutation. Science Neurodegenerative Diseases (Calne, D.B., ed.), tangles in Alzheimer’s disease. Ann. Neurol. 19 Duyao, M. et al. (1993) Trinucleotide repeat 35 Markesbery, W.R. (1997) Oxidative stress 2 Fearnley, J.M. and Lees, A.J. (1991) Ageing and length instability and age of onset in Huntington’s hypothesis in Alzheimer’s disease. Free Rad. Biol. Parkinson’s disease: substantia nigra regional selectivity. Brain 114, 2283–2301 20 Davies, S.W. et al. (1997) Formation of neuronal 36 Browne, S. et al. (1999) Oxidative stress in 3 Schiffer, D. et al. (1991) Ubiquitin in motor neuron intranuclear inclusions underlies the neurological Huntington’s disease. Brain Pathol. 9, 147–163 disease: Study at the light and electron microscope.
dysfunction in mice transgenic for the HD 37 Choi, D.W. (1988) Glutamate neurotoxicity and J. Neuropathol. Exp. Neurol. 50, 463–473 diseases of the nervous system. Neuron 1, 623–634 4 Majno, G. and Jois, I. (1995) Apoptosis, oncosis, 21 Sadou, F. et al. (1998) Huntingtin acts in the 38 Bergeron, C. (1995) Oxidative stress: its role in and necrosis. An overview of cell death. Amer. nucleus to induce apoptosis but death does not the pathogenesis of amyotrophic lateral sclerosis.
correlate with the formation of intranuclear J. Neurol. Sci. 129 (suppl) 81–84 5 Selkoe, D.J. (1998) The cell biology of α-amyloid 39 Rothstein, J. et al. (1990) Abnormal excitatory protein and presenilin in Alzheimer’s disease.
22 Shimohata,T. et al. (2000) Expanded amino acid metabolism in amyotrophic lateral polyglutamine stretches interact with TAFII130, sclerosis. Ann. Neurol. 28, 18–25 6 Esler, W.P. et al. (2000) Transition-state analogue interfering with CREB-dependent transcription.
40 Lin, C.L.G. et al. (1998) Aberrant RNA processing inhibitors of γ-secretase bind directly to in a neurodegenerative disease: the cause for presenilin-1. Nat. Cell Biol. 2, 428–434 23 Mehraein, P. et al. (1975) Quantitative study on absent EAAT2, a glutamate transporter, in 7 Sherrington, R. et al. (1995) Cloning of a gene dendrites and dendritic spines in Alzheimer’s amyotrophic lateral sclerosis. Neuron 20, 589–602 bearing missense mutations in early-onset disease and senile dementia. Adv. Neurol. 41 Takuma, H. et al. (1999) Reduction of GluR2 RNA familial Alzheimer’s disease. Nature 375, 754–760 editing, a molecular change that increases calcium 8 Yankner, B.A. (1996) Mechanisms of neuronal 24 Davies, C.A. et al. (1987) A quantitative influx through AMPA receptors, selective in the degeneration in Alzheimer’s disease. Neuron morphometric analysis of the neuronal and spinal ventral gray of patients with amyotrophic synaptic content of the frontal and temporal lateral sclerosis. Ann. Neurol. 46, 806–815 9 Spillantini, M. et al. (1998) Tau protein pathology cortex in patients with Alzheimer’s disease.
42 Zuo, J. et al. (1997) Neurodegeneration in Lurcher in neurodegenerative diseases. Trends Neurosci. mice caused by mutation in ∆ 2 glutamate 25 Klaue, R. (1940) Parkinsonsche Krankheit receptor gene. Nature 388, 769–773 10 Mann, D.M. (1989) Cerebral amyloidosis, aging (Paralysis agitans) und postencephalitischer 43 Luthi-Carter. R. et al. (2000) Decreased and Alzheimer’s disease; a contribution from expression of striatal signaling genes in a mouse studies on Down’s syndrome. Neurobiol. Aging klinischanatomischen Differentialdiagnose. Arch. model of Huntington’s disease. Hum. Mol. Genet. Psychiat. Nervenkrankh. 111, 251–321 11 Takashima, A. et al. (1993) Tau protein kinase I is 26 Anglade, P. et al. (1997) Apoptosis and autophagy 44 Kastner, A. et al. (1993) Tyrosine hydroxylase in nigral neurons of patients with Parkinson’s protein and messenger RNA in the dopaminergic neurotoxicity. Proc. Natl. Acad. Sci. U. S. A. disease. Histol. Histopathol. 12, 25–31 nigral neurons of patients with Parkinson’s 27 Kiernan, J.A. and Hudson, A.J. (1993) Changes in disease. Brain Res. 606, 341–345 12 Polymeropoulos. M.H. et al. (1997) Mutation in shapes of surviving motor neurons in amyotrophic 45 Yamamoto, A. et al. (2000) Reversal of the α-synuclein gene identified in families with lateral sclerosis. Brain 116, 203–215 neuropathology and motor dysfunction in a Parkinson’s disease. Science 276, 2045–2048 28 Roos, R.A. and Bots, G.T. (1983) Nuclear conditional model of Huntington’s disease. Cell membrane indentations in Huntington’s disease.
Drosophila model of Parkinson’s disease. Nature 46 Schenk, D. et al. (1999) Immunization with amyloid-β 29 Finch, C.E. and Day, J.R. (1994) Molecular biology attenuates Alzheimer’s disease-like pathology in the 14 Jensen, P.H. et al. (1998) Binding of α-synuclein of aging in the nervous system: a synopsis of the PDAPP mouse. Nature 400, 173–177 to brain vesicles is abolished by familial levels of mechanisms. In Neurodegenerative 47 Morgan, D. et al. (2000) Aβ peptide vaccination Parkinson’s disease mutation. J. Biol. Chem. Diseases (Calne, D.B., ed.), pp. 33–50, Harcourt prevents memory loss in an animal model of 30 Brody, H. (1980) The nervous system and aging.
Alzheimer’s disease. Nature 408, 982–985 15 Shimada, H. et al. (2000) Familial parkinson 48 Björklund, A. and Lindvall, O. (2000) Parkinson 31 Montine, T.J. et al. (1996) 4-Hydroxy-2-nonenal is disease gene therapy moves toward the clinic.
ubiquitin-protein ligase. Nat. Genet. 25, 302–305 cytotoxic and crosslinks cytoskeletal proteins in 16 Hirano, A. et al. (1984) Fine structural P19 neuroglial cultures. Am. J. Pathol. 148, 89–93 49 Freeman, T.B. et al. (2000) Transplanted fetal observations of neurofilamentous changes in 32 Munch, G. et al. (1997) Advanced glycation striatum in Huntington’s disease: phenotypic amyotrophic lateral sclerosis. J. Neuropathol. endproducts in aging and Alzheimer’s disease.
development and lack of pathology. Proc. Natl. Acad. Sci. U. S. A. 97, 13877–13882 17 Deng, H.X. et al. (1993) Amyotrophic lateral 33 Beal, F. (2000) Energetics in the pathogenesis of 50 Kawasaki, H. et al. (2000) Induction of midbrain sclerosis and structural defects in Cu–Zn neurodegenerative diseases. Trends Neurosci. dopaminergic neurons from ES cells by stromal superoxide dismutase. Science 261, 1047–1051 cell-derived inducing activity. Neuron 28, 31–49 Letters to the Editor
Trends in Molecular Medicine: a forum for comment and debate The Editor welcomes correspondence from readers of Trends in Molecular Medicine on any articles published. Please stateclearly whether or not you wish the letter to be considered for publication. If a letter is considered suitable for publication, itwill be sent to the author of the original article for their response; both the letter and reply will be published together.
Please note: submission does not guarantee publication.

Source: http://neurofuture.ru/_mozg/archive/content/nov11Guliaeva/How%20do%20neurons%20die.pdf