James C. Vickers, Carolyn E. King, Tracey C. Dickson, Paul A. Adlard, Helen L. Saunders and Irene Jacobs

Neurobiology Laboratory, Division of Pathology, Clinical School, University of Tasmania, 43 Collins Street, Hobart, Tasmania 7000.

How do plaques cause the neuronal degeneration of Alzheimer's disease?

Several ideas about the link between ß-amyloid plaque formation and nerve cell death have been previously put forward.

• ß-amyloid is toxic.
• A plaque-associated molecule is toxic.
• ß-amyloid potentiates the toxicity of other molecules.
• ß-amyloid deposition promotes an inflammatory reaction that is harmful to neurons.

We have recently suggested that ß-amyloid deposition is similar to other amyloidegenic diseases in that the damage it causes surrounding tissue is physical in nature, and it is the reaction of nerve cells to this microscopic trauma over a long period of time that leads to the neurofibrillary pathology of Alzheimer's disease (Vickers, 1997; King et al., 1997).

Clues to the origin of neurofibrillary pathology.

Abnormal (dystrophic) neurites associated with ß-amyloid deposits in Alzheimer's disease can be divided into three types based on their immunolabelling profile with antibodies to cytoskeletal proteins such as tau and neurofilaments (Masliah et al., 1993; Vickers et al., 1994; Su et al 1996).

• Dystrophic neurites labelled excusively with antibodies to tau.
• Dystrophic neurites labelled excusively with antibodies to neurofilaments.
• Dystrophic neurites with an outer rim labelled for neurofilaments and an inner core immunoreactive for tau (and stained with thioflavine S).

This has led to the hypothesis that dystrophic neurites 'mature' from forms containing accumulations of neurofilaments to the structures labelled for tau and stained with thioflavine S.

The earliest neuronal alterations associated with Alzheimer's disease.

We examined aged individuals (n=8) who demonstrated ß-amyloid plaques in their neocortex but were not demented and lacked the widespread neuronal degeneration of Alzheimer's disease. These cases are likely to represent the preclinical stage of the disease (Morris et al., 1996).

In preclinical Alzheimer's disease cases, double immunofluorescence immunolabelling showed that ß-amyloid immunoreactive plaques of the association neocortices (particularly those plaques in layers II and III) were associated with abnormal neurites containing accumulations of neurofilaments (arrows). These abnormal neurites had either a ring- or bulb-like morphology. A neurofilament immunoreactive neuron is shown on the right.

These early dystrophic neurites lacked tau immunolabelling and were not stained with thioflavine S.The abnormal neurites resembled axons responding to physical damage, a stereotyped reaction that intimately involves specific changes to neurofilaments.

A rodent model of physical injury mimics the early neuronal pathology of Alzheimer's disease.

Following on from our observation of the early forms of dystrophic neurites in Alzheimer's disease, we hypothesised that physical damage to axons in a model system may replicate this initial neuronal pathology. To test this, we placed a 25 gauge needle into the somatosensory cortex of Wistar rats anaesthetised with sodium pentobarbitol (60 mg/kg i.p.) and immobilised in a stereotaxic frame. Twenty four hours later, animals were re-anesthetised and perfused with 4% paraformaldehyde. Brain sections were then immunoreacted with the antibodies to neurofilament proteins used in the human studies.

These studies confirmed that physical damage produces the ring and bulb-like accumulations of neurofilaments. Similar neurofilament-containing, palque-associated abnormal neurites have been reported to occur in transgenic mice models of ß-amyloid plaque formation (Masliah et al. 1996).

Summary

The earliest neuronal alterations associated with ß-amyloid plaque formation resemble physically damaged axons, and neurofilaments are crucially involved in this reaction of nerve cells to trauma.

This data has led to the proposal of a mechanism linking the formation of ß-amyloid deposits with the subsequent development of neurofibrillary pathology and nerve cell degeneration. Abnormal neuritic sprouting and neurofibrillary degeneration may be due to the persistent stimulation of the programmed response to physical damage. Neurofibrillary tangles are likely to occur in the cell bodies of origin of damaged axons due to the over-activation of the retrograde reaction.

This new hypotheis may also explain why head trauma is a risk factor for Alzheimer's disease, as previous damage to a neuron "primes" it for an exaggerated response to subsequent damage (Vickers, 1997). Similarly, repeated head trauma (e.g. in boxers or individuals who show repeated self head injury (Hof et al., 1990, 1992)) may also be sufficient to produce the neuronal damage response that eventually leads to neurofibrillary pathology.

References

Hof et al. (1990) Acta Neuropath. 82:321-326.

Hof et al. (1992) Acta Neuropath. 85: 23-30.

King et al. (1997) Neuroreport 8: 1663-1665.

Masliah et al. (1993) Am. J. Path. 142: 871-882.

Masliah et al. (1996) J. Neurosci. 16: 5795-5811.

Morris et al. (1996) Neurology 46: 707-719.

Su et al. (1996) Brain Res. 739: 79-87.

Vickers et al. (1994) Neuroscience 62: 1-13

Vickers, J.C. (1997) Neuroscience 78: 629-639.

Acknowledgments

Funded by the Australian National Health and Medical Research Council (NH&MRC), Department of Veterans Affairs, Tasmanian Acute Care Program, Ramaciotti Foundation, Ian Potter Foundation, AMRAD Corp., MSD Foundation and the Tasmanian Masonic Centenary Medical Research Foundation. We would like to thank the NH&MRC Brain Bank (Adelaide) and the Department of Pathology (University of Sydney) for the provision of human material. Thanks also to Dr Greg Woods for help in html instruction. The homepage of the Division of Pathology, Univesity of Tasmania is here. James Vickers can be contacted by e-mail.