Year: 2006
Type: article
Abstract: “When we hear news we should always wait for the sacrament of confirmation.” —Voltaire (1694–1778) In this issue of the Annals, a series of point–counterpoint articles are presented that follow up the report by Kevin McNaught and colleagues nearly 2 years ago.1 The McNaught article, among the most highly cited articles published in the Annals in 2004, continues to attract attention and debate. McNaught and colleagues described a novel model for Parkinson's disease (PD) induced in Sprague-Dawley rats by the systemic administration of a proteosome inhibitor. Noteworthy features of the model were the development of a clinically obvious extrapyramidal syndrome, selective loss of nigrostriatal neurons, intracytoplasmic aggregates resembling Lewy bodies, and a response to dopamine agonists. The model was most exciting in part because the concept underlying the experiments was biologically sound; that is, genetic and epidemiological studies of PD support the importance of proteosome dysfunction in the pathogenesis of PD, and prior work had shown vulnerability of cultured dopaminergic neurons to proteosome inhibitors.1, 2 Thus, one begins with the presumption that the plausibility of the findings was reasonably high in the light of current thinking about the biology of PD. In this issue, two independent groups report that they were able to replicate, at least in part, several features of the McNaught model3, 4; however, the model could not be replicated in any manner by at least three other laboratories.5-7 In an accompanying commentary, Flint Beal and Tony Lang2 attempt to reconcile these discordant findings, and they conclude quite reasonably that, given the current state of knowledge, the model should not be considered validated until additional confirmatory studies are available. The authors of the original article are also to be commended for initiating a process that should resolve the remaining ambiguities and discrepancies.8 For numerous animal models of human disease, variability of findings are, exasperatingly, observed far more commonly than generally appreciated. The sources of variability are many and may include genetic differences (drift can occur even in inbred strains), epigenetic differences, minor changes to protocols including batch variability of reagents, and environmental variables related to housing, temperature, season, diet, infections, and stress. It is noteworthy that variability and inconsistency across, and even within, laboratories has been observed similarly in other toxin-induced models for PD, specifically those mediated by 6–hydroxydopamine, MPTP, the insecticide rotenone, and the herbicide paraquet, and in genetic models mediated by overexpression of α-synuclein in Drosophila and in mice.9, 10 Science does not proceed in a straight line. Initial observations usually report effect sizes of greater magnitude than found in follow-up or confirmatory studies.11 Potential biases exist at many levels across the process by which scientific information is communicated.12 The responsibility of the Annals does not end when an accepted manuscript has been published but must continue throughout the life of the discovery, including the need to consider serious inquiries, challenges, and other significant developments. We hope we have carried out this obligation by fully airing the proteosome inhibition model in this issue.
Source: Annals of Neurology
DOI: 10.1002/ana.20941
Author: Stephen L. Hauser
Institution: University of California, San Francisco
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2026215117