Introduction

Parkinson's disease is, after Alzheimer's disease, the commonest human neurodegenerative illness but its cause or causes remains largely unknown, save for a few uncommon genetic varieties. Over the years, many clinical and experimental models have been developed attempting to elucidate its pathogenesis, with the hope of more effective treatments.1 The models relating to its pathogenesis fall into three groups: genetic, studied in transgenic mice, pharmacological, such as reserpine, and toxic, first imputed many years ago.

Interest in the toxic group dates to at least 1882, with Pölchen's description of a parkinsonian syndrome with symmetrical pallidal lesions after ‘Kohlendunst’ (coal gas) poisoning.2 Edsall and Drinker (1919), contributing to a Festschrift for Sir William Osler, detailed the clinical features of chronic manganese poisoning after inhalation, ingestion and exposure to industrial pollution3; Edelman (1921) described similar brain changes after gaseous hydrocyanic acid.4 In 1930, others followed: Buzzo and Guerra after potassium cyanide,5 Negro after carbon disulphide6 and Haurowitz, who appears to be the first to demonstrate focal deposits of copper in what we now call Wilson's disease.7

Today the MPTP (1-methyl-phenyl-1,2,3,6 tetrahydropyridine) model is probably the best known. In 1982, a group of young drug addicts developed an illness closely resembling Parkinson's disease after self-administrating a contaminated synthetic heroin analogue.8 MPTP produced many but not all of the clinical and pathological changes of the idiopathic disease in monkeys and degeneration of dopaminergic neurons in mice but not in rats. Other toxins have included paraquat, a herbicide;9 rotenone, a natural occurring cytotoxic drug employed as an insect and fish poison;10 and isoquinoline derivatives. …