Subscribe to eTOC

Is Parkinson's Disease an Autoimmune Disorder?

Article In Brief

Findings from a new study suggest that environmental factors—specifically those in the gut—may increase susceptibility to Parkinson's disease and that mitochondrial quality control, mediated by the lysosome, plays a central role in the disease.

PINK1 is an extremely rare cause of Parkinson's disease (PD), and mouse models bearing homozygous PINK1 mutations don't develop a movement disorder, which has made the gene's relevance to understanding the human disease seem somewhat tenuous. But a new study in the journal Nature may bring PINK1 closer to the center of interest in PD, because it shows that PINK1 mice can indeed develop PD-like symptoms, but only if they are first infected with a common gut bacterium.

The results demonstrate an important immune-related function of PINK1, suggest that human PINK1 PD may be a mitochondria-related autoimmune disease, and highlight the potential role of the environment in causing neurologic disease.

“The evidence is getting stronger that an autoimmune response may be involved in Parkinson's disease,” said David Sulzer, PhD, of Columbia University, who was not involved in the study. “This paper is an additional reason to consider that the immune system may be playing a role in the disease.”

The idea that the immune system might contribute to PD is not new, but the centrality and extent of its contribution has remained controversial. Mitochondria too have long been a focus of pathogenic hypotheses in PD, a focus that became more prominent with the discovery of two PD genes: parkin and PINK1. Parkin's function in flagging damaged mitochondria for degradation by the lysosome, a process known as mitophagy, has been clear, and PINK1 has also been thought to play a role in mitophagy.

At the lysosome, mitochondrial antigens are processed and then shipped to the cell surface, where they are presented to the immune system in MHC class I molecules. That led Michel Desjardins, PhD, and Louis-Eric Trudeau, PhD, to ask what would happen if PINK1 was absent.

“PINK is required for mitophagy, so in principle if we were to knock out PINK, there would be no antigen presentation of mitochondrial proteins,” said Dr. Desjardins, professor of pathology and cell biology in the School of Medicine at the University of Montreal. But in a 2016 paper in the journal Cell, they showed that knocking out PINK stimulated mitochondrial antigen presentation, especially in the context of lipopolysaccharide (LPS), the major component of the membrane of Gram-negative bacteria, and a standard laboratory trigger for immune presentation.

The Next Logical Experiment

That led the team to the next logical experiment, whose results were published in the new study, of subjecting PINK1-knockout mice to bacteria themselves, not just LPS.

They exposed both wild-type and PINK1-knockout mice to Citrobacter rodentium, a mouse intestinal pathogen that is used to model human infection with enteropathogenic E. coli. The infection course was similar in both mice, with a peak of colonization at 13 days and resolution by 20 days. Levels of fecal markers of inflammation were similar between the two, as were circulating cytokines and other signs of inflammation. Both mice developed immunity to further infection as well.

What differed between wild-type and PINK1-knockout mice was that in the knockout, mitochondrial antigens were presented to the immune system, and this presentation led to the establishment of a clone of CD8+ killer T cells that responded to the antigens. The team showed that these peripherally generated T cells crossed into the mouse brain.

For these killer T cells to cause damage in the brain, the authors hypothesized, they would need to encounter mitochondrial antigens on the surface of brain tissue, triggering an autoimmune attack. To explore this question, they turned to cell culture. They showed that in mixed neuronal culture from substantia nigra of PINK1-knockout mice, LPS stimulated the production of MHC class I molecules and presentation of mitochondrial antigens, followed by the death of large numbers of dopaminergic neurons.

That result suggested that infection might cause a dopaminergic deficit in PINK1-knockout mice. The team confirmed that hypothesis when they observed, beginning four months post-infection, that even though the mice had recovered from their acute illness, they developed slowed movements, difficulty lifting their head, and a reduction in hind limb activity. Affected mice performed poorly on the pole test, during which the mouse must descend a thin pole, but their performance was normalized by administration of levodopa. Importantly, in the mice, unlike in culture, dopamine neurons were dysfunctional but not dead, with shrunken processes but living cell bodies.

The conclusion, according to Drs. Desjardins and Trudeau, is that a central function of PINK1 is to inhibit presentation of mitochondrial antigens, and that in its absence, and in the presence of a bacterial challenge, those antigens lead to a T-cell mediated autoimmune reaction against dopamine-producing cells in the brain. “In the absence of PINK, when you stress the animal” with LPS or Gram-negative bacteria, “you stimulate an autoimmune response,” said Dr. Desjardins. “PINK1 is an immune system mediator.”

Unlike the irreversible progression of PD, though, the movement disorder in the PINK1-knockout mouse was transient, and by 12 months, the mice had returned to normal, with full restoration of their motor ability.

“At first this was surprising,” Dr. Desjardins said, “but as we thought about it, it began to fit. We infected the mice with only one type of bacteria, and they developed a motor impairment, but they were protected against reinfection.” The initial infection apparently caused dysfunction, rather than the death, of dopamine neurons, allowing a restoration of function as the neurons regrew their processes.

“In humans it is very unlikely that an individual will be infected by only one type of bacteria over the lifespan,” which may lead to chronic dysfunction and eventual loss of neurons, he said. They will be testing this hypothesis in mice by reinfecting with different species sequentially.

All cells contain mitochondria, so why are dopamine neurons especially susceptible to this autoimmune response? “This is a global question in Parkinson's disease,” said Dr. Trudeau, who is a professor of pharmacology and physiology at the University of Montreal. One clue is provided by a publication from the lab of Dr. Sulzer, of Columbia, who showed that dopamine neurons appear to express more MHC class I than other neuronal types. “When dopamine neurons are stimulated, they may do more antigen presentation,” Dr. Trudeau said. “This is one hypothesis, and we are testing it.”

The findings from this study support two major ideas in PD pathogenesis, Dr. Desjardins said. First is that environmental influences—specifically those in the gut—may increase susceptibility to PD. Whether there are infectious contributors and autoimmune mediation in other forms of PD remains to be seen, but even if not, these results focus attention on the second major idea: that mitochondrial quality control, mediated by the lysosome, plays a central role in the disease.

“When dopamine neurons are stimulated, they may do more antigen presentation. This is one hypothesis, and we are testing it.”


“The majority of the genes for the genetic forms of Parkinson's disease code for proteins that act in either the mitochondria or the lysosome,” Dr. Desjardins said. “We think there is a strong possibility that all of these PD genes will affect the mitochondrial antigen presentation pathway to a certain extent,” either by increasing presentation, as with PINK1, or impairing lysosomal function, as with other PD genes.

Expert Commentary

“This study provides additional reasons to consider that autoimmunity may contribute to Parkinson's disease,” said Dr. Sulzer, professor of neurobiology in the department of neurology at Columbia University. “These results suggest that, under some conditions, T cells of the immune system could could directly interact with the substantia nigra neurons that die in PD. To me, that's very exciting.” Nonetheless, he noted, “This is by no means a full model of PD. There are some aspects that are missing,” including progressive neurodegeneration. “This could be a very important part of the jigsaw puzzle, but there are still pieces of the jigsaw puzzle that are missing, that we still don't understand.”


Drs. Desjardins, Trudeau, and Sulzer reported no competing interests.

Link Up for More Information

• Matheoud D, Cannon T, Voisin A, et al. Intestinal infection triggers Parkinson's disease-like symptoms in Pink1-/- mice Nature 2019:571(7766):565–569.
• Matheoud D, Sugiura A, Bellemare-Pelletier A, et al. Parkinson's disease-related proteins PINK1 and parkin repress mitochondrial antigen presentation Cell 2016;166(2):314–327.