Somatic mutation: a cause of sporadic neurodegenerative diseases?

Abstract

The major causes of the common neurodegenerative diseases remain unknown. Alzheimer’s disease, Parkinson’s disease and motor neuron disease occur in both sporadic and familial forms, and mutations are progressively being found in families with these disorders. However, attempts to find causative mutations in blood DNA from the sporadic forms of the diseases have proved fruitless. It is hypothesised that this is because the causative mutations are found only in the cells in the central nervous system that are affected by the disease. These mutations arise in the developing embryo in progenitor cells of neurons or glia. The diseases are not passed to offspring since the mutations are not present in the germ-line. To find somatic mutations, the affected central nervous system cells need to be separated out and submitted to DNA analysis.

Introduction

The major neurodegenerative disorders, Alzheimer’s disease, Parkinson’s disease and motor neuron disease (MND) have one feature in common: each has a familial and a remarkably similar sporadic form. The familial forms are less common and are inherited via germ-line mutations. From the point of view of mutation detection the familial forms are “low-hanging fruit” since the technology is in place to find the altered DNA, given a large enough family. A greater challenge is to find the causes of the more common sporadic neurodegenerative diseases, a group in which neither parents nor offspring of the case are affected. For the sake of brevity most of the following examples given to support the present hypothesis refer to MND, though similar examples exist for Alzheimer’s and Parkinson’s disease.

Two predominant hypotheses have arisen concerning the causes of the sporadic neurodegenerative disorders. The first is that these disorders are due to environmental agents, a concept that fits with the conditions affecting only one family member. Attempts continue to be made to find environmental agents that cause neurodegenerative disorders. For example, enteroviruses have often been implicated in MND [1]. However, a feature that makes an environmental cause unlikely is the very similar appearance of the familial and sporadic forms of the neurodegenerations. A clinician is not able to distinguish between familial and sporadic MND in an individual patient and a pathologist cannot reliably differentiate between the two conditions on tissue examination. This similarity suggests that genetic mechanisms underlie the sporadic conditions as well. Another argument against an environmental agent is the late age of onset of sporadic neurodegenerative conditions. Why should a viral infection of motor neurons manifest only later in life as MND? Thirdly, epidemiological studies have not found any strong associations between environmental factors and neurodegenerative diseases. Although agents such as agricultural chemicals have been linked with MND, the associations are weak and inconsistent [2].

Reservations about environmental causes of neurodegenerations have led many to suggest that a genetic susceptibility to an environmental agent underlies Alzheimer’s disease [3], Parkinson’s disease [4] and MND [5]. For example, a link between pesticide exposure and polymorphisms in genes responsible for pesticide detoxification has repeatedly been looked for in Parkinson’s disease [6]. A variation on this theme is the gene–gene hypothesis where two or more genes need to be mutated to cause the condition, or combinations of a number of disordered genes with a number of environmental agents. A weakness in these gene-environment hypotheses is that the sporadic neurodegenerations are for the most part truly sporadic. For example, even when large families are studied, neither close nor distant relatives of people with sporadic MND have the condition. If genetic susceptibility played a part in the disease, one would expect family members to have at least a slightly increased risk of getting the disease since some family members would carry the same genetic susceptibility and would be exposed to the same environmental agent. For the more common neurodegenerations such as Parkinson’s disease the evidence for isolated sporadic cases is admittedly less clear-cut, with some reports of an increased incidence of the disease in relatives with late-onset Parkinson’s disease [7].

Unlike germ-line mutations that affect every cell in the body, somatic mutations involve only a particular cell or cell line [8]. This leads to mosaicism, the presence of more than one genetically distinct cell in a single organism (Fig. 1). Somatic mutations account for normal antigenic diversity, are involved in the pathogenesis of mitochondrial disorders, underlie most neoplasms and can explain why monozygotic twins are not identical but show phenotypic variability [8]. These mutations arise spontaneously or from external agents such as irradiation or toxins [9]. Most research into somatic mutations has been in the field of cancer, but it is becoming apparent that other disorders may be caused by these mutations [10]. However, the concept of somatic mutations as a cause of the common neurodegenerative disorders does not seem to have previously been considered.

Developing tissue harbours somatic mutations at least as frequently as adult tissue. In mice about half of all somatic mutations arise during development and half of these are in utero [11]. In humans, somatic mutations in stem cells that differentiate into B and myeloid cells are found in leukaemia [12]. In addition to somatic gene mutations, somatic chromosomal abnormalities can cause alterations in gene expression and can lead to mosaicism. In the developing nervous system of the mouse about one-third of neuroblasts have chromosomal abnormalities [13]. It is therefore becoming apparent that developing cells in the CNS are susceptible to a range of DNA and chromosomal alterations.

Animal experiments indicate that most somatic tissues, whether quiescent or actively proliferating, have similar frequencies of gene mutation [11]. In the adult mouse brain about 1% of cerebral cortical neurons are aneuploid in the XY chromosomes alone, hence the total number of adult neurons with chromosomal abnormalities is likely to be higher [13]. Non-dividing mature neurons can be expected to contain as many somatic gene or chromosomal abnormalities as dividing cells.