Chronic neurodegenerative disease is characterized by extensive regional loss of neurons in the brain and neuropathological hallmarks in surviving neurones. Genetic modelling by overexpression of hallmark proteins does not produce extensive neurodegeneration, whereas genetic deletion of neuronal 26S proteasomes does, as well as some hallmarks of human disease.
Chronic neurodegenerative diseases including Alzheimer's disease, dementia with Lewy bodies and Parkinson's disease are characterized by the neuropathological features of deposits of so-called amyloidogenic proteins, including intraneuronal inclusions containing proteins such as hyperphosphorylated tau and α-synuclein and extraneuronal deposits of fragments of the Alzheimer precursor protein. Additionally, there is extensive regional neurodegeneration-neuronal death-in all the diseases [1,2].
It is a curious fact that some 14 years of genetic modelling of chronic neurodegenerative disease by overexpression of normal or mutated versions of human amyloidogenic proteins has not produced widespread regional neurodegeneration, in spite of producing amyloid plaques and intraneuronal deposits of proteins, but not the intraneuronal inclusions characteristic of human disease! Most attempts at modelling disease have relied on mouse [3–5] or fruitfly  transgenesis. In the mouse there is scant evidence of neurodegeneration in regions afflicted in man: in the fruitfly there is evidence of neurodegeneration, particularly in the eye. There is no obvious reason in the mammal for lack of neurodegeneration but the mice are often ‘blamed’ for not cooperating, possibly because of their short lifespan [3–5]!
In humans, there is little knowledge of the relationship between regional neuronal death and the biogenesis of intraneuronal and extraneuronal amyloid deposits. Molecular neuropathological approaches are limited because of the obvious limitations of autopsy material. Since multiple biopsies are generally ethically not permitted, disease progression in the individual brain cannot be followed. Therefore it is not clear as to whether plaques and tangles trigger neuronal death and accompanying gliosis or whether neurons dying for other molecular reasons generate the neuropathological hallmarks of disease. The neuropathological investigations describe the situation in the brain that remains at death and does not show what happened to the neurons that have died; these cells are gone. The relationship between gliosis and neuronal death is similarly not clear. Gliosis may be a morphological response to neuronal demise or glial cells may proliferate to produce trophic factors that either promote neuronal death or protect neurons from death processes as well as triggering remodelling of neuronal connections to preserve vestiges of neuronal networks to preserve cognition and memory. In amyotrophic lateral sclerosis, non-motor neuronal autonomous contributions to cell death are supported by experimental evidence . In this condition, evidence also exists that muscle cells can exert molecular cell survival influences on innervating motor neurons .
One difficulty is that the mechanism(s) of neuronal death are not well established. Although evidence for apoptopic processes have been documented, the precise events in the death of human brain neurons are not known . For example, whether neuronal death is preceded by extensive dye-back of neuritic processes is unclear. The time taken for neurons to die is not known or whether neurons devoid of extensive connections can survive for prolonged periods. The clearance of dead neurons by glial cells is not fully characterized. There are also immunological and inflammatory responses involved in the elimination of dying neurons. The extent of compensatory neuronal reconnections involving synaptic plasticity to slow or preserve memory is not understood. The degree to which memory can be preserved by such compensations before the onset of clinical symptoms is also not known.
Neuronal survival is almost certainly orchestrated by complex intercellular neuron–neuron and neuron–glial trophic signalling mechanisms as well as adequate vasculature with oxygen and nutrient supply. Neurons seem to eventually respond by tipping the balance between neuronal survival and death mechanisms in favour of neuronal death. These cellular mechanisms are not understood and may involve apoptotic, necrotic and necroapoptotic processes . The neuron-specific autonomous processes will eventually decide whether a specific neuron will survive or die? The death of single neurons may have consequences for numerous other neurons, triggering a regional death cascade in the brain.
Set against these general cell biological considerations is the fact that mutations in genes for the Alzheimer precursor protein, presenillins, tau, α-synuclein, progranulins, huntingtin, TDP43 (TAR DNA-binding protein 43) and Fus predispose to familial chronic neurodegenerative disease. Based on these findings there has been an enormous amount of work in cell models and transgenic animals attempting to delineate molecular pathways that could connect functional biochemical pathways related to the functions and malfunctions of these molecules to neuronal death. For Alzheimer's disease, much of this work has focused on the so-called amyloid cascade hypothesis. The notion has lead to the promulgation of the idea that removal of extracellular and intraneuronal amyloid deposits may offer respite from disease. This theory has been taken to the point of immunizing against Aβ (amyloid β-peptide) to slow down or prevent disease. The initial clinical trials had to be stopped because of neuroinflammation and patient death. Neuropathological investigation showed some removal of extraneuronal amyloid deposits but there was no robust statistical evidence for cognitive improvement in patients, allowing for the fact that treatment had to be terminated .
Of course, it needs to be remembered that there is the counter-view that intraneuronal inclusions containing α-synuclein, tau, huntingtin, TDP43 and Fus and extraneuronal Aβ deposits may be deliberately formed as a protective mechanism in the brain to slow down or prevent neurodegeneration, e.g. . In this case, therapeutic removal of these aggregates may be contraindicated in the brain. As indicated previously, overexpression of these proteins in mammalian brains does not cause neurodegeneration to any great extent in spite of giving rise to some intraneuronal aggregates and particularly extra-neuronal protein deposits. Maybe this is because the excess of these proteins is rendered less toxic by the formation of intraneuronal inclusions or extraneuronal aggregates. With the exception of APP (amyloid precursor protein)-bearing chromosome 21 duplication in Down's syndrome and α-synuclein gene replication, there is no evidence for the overproduction of amyloidogenic proteins in the brain. It is likely that the brain response to overproduction of these proteins is to form the aggregates.
It is apodictic that patients with mutations in Alzheimer-related amyloidogenic proteins develop neurodegeneration, often with much earlier onset than patients with idiopathic disease. This is also true for other ‘monogenic’ mutations in intraneuronal proteins, e.g. in the huntingtin gene. It is not surprising that the continuous production of a mutated amyloidogenic protein will trigger a pathological response as also evidenced outside the nervous system by muscular dystrophy with dystrophin mutations, pancreatic disorder with α1-anti trypsin mutations and lung disease with mutated transmembrane conductance regulator in cystic fibrosis. The genetic mutations in chronic human familial neurodegenerative disease may give rise to nosologically distinct diseases from idiopathic disorders even with amyloid deposits in the latter. Neurons dying in the brain may generate and accumulate extraneuronal fragments of APP and intercellular deposits of tau, α-synuclein, etc.
If this notion is correct, then what sort of neuron autonomous perturbations and malfunctions (and consequential non-neuronal responses, e.g. microglial activities) could cause such extensive neuronal demise as seen in human disease? A feature of all chronic neurodegenerative diseases is ubiquitination of proteins in intraneuronal inclusions. The diversity of the functions of protein ubiquitination in the cell, to rival protein phosphorylation, suggests that ubiquitin-based signalling responses and ubiquitin-dependent protein degradation by the 26S proteasomal and selective autophagic systems are central to neuronal homoeostasis . Age-related malfunction of these systems, in predominately non-dividing neurons, may cause the formation of inclusions containing ubiquitylated proteins and trigger neurodegeneration. Experimentally, genetic ablation of a 26S proteasome regulatory ATPase gene and autophagy genes in the brain causes neurodegeneration [14–16]. While it would be inadvisable to claim that protein catabolic deficiencies are the only cause of human neurodegeneration, it is clear from mammalian genetic models that the inability to degrade proteins in brain neurons causes cell death (and inclusions resembling human Lewy bodies with 26S proteasome depletion) in contrast with the overexpression of so-called amyloidogenic proteins.
Models of Dementia: the Good, the Bad and the Future: A Biochemical Society Focused Meeting held at Robinson College, Cambridge, U.K., 15–17 December 2010. Organized and Edited by Stuart Allan (Manchester, U.K.), Christian Hölscher (University of Ulster, Coleraine, U.K.), Karen Horsburgh (Edinburgh, U.K.), Simon Lovestone (King's College London, U.K.) and Calum Sutherland (Dundee, U.K.).
Some of the work reported here was supported by the Pathological Society of Great Britain and Ireland, Wellcome Trust, BBSRC, Alzheimer's Disease Society, Neuroscience Support Group at the Queen's Medical Centre and Parkinson's U.K.