Skip to main content

Biological News

Parkinson’s As An Autoimmune Disease: More Evidence

For many complex diseases, you’ll find that there are a couple of hypotheses floating around them that are hard to prove and hard to disprove: one is that they’re actually caused by some (as yet unrecognized) infectious agent, and the other is that that they’re actually an autoimmune/inflammatory disorder. You can also recognize that these two can have features in common, as seen in something like Guillian-Barré syndrome, where a (usually innocuous and often hardly noticed) viral infection or other stimulus can lead to a sudden autoimmune crisis. A whole list of conditions have had such explanations attached to them, more or less persuasively: Alzheimer’s, obesity, various forms of arthritis (with little doubt on the autoimmune side), diabetes (Type I, certainly, but even Type II), multiple sclerosis, Parkinson’s, and more. Those links lead mainly to autoimmune explanations, but infectious-agent hypotheses are found quite easily as well, and going back many years.

A new paper adds what might be strong evidence to the Parkinson’s explanation. It’s been known for some time that there’s an association between the disease and MHC (major histocompatibility complex) alleles although (at the same time) having another autoimmune disease doesn’t seem to raise the risk for Parkinson’s itself. That’s interesting, in that the brain has mostly been thought of as an “immunoprivileged” compartment, but it’s also been increasingly clear that this doctrine is not as solid as it might be. Many CNS conditions have an undeniable inflammatory component, and it’s also quite possible that they lead to less-restrictive blood-brain barriers (or perhaps it’s the latter leading to the former?) In 2014, a paper came out (from some of the same authors on the current one) on MHC expression in postmortem neuronal tissue from Parkinson’s patients suggesting that catecholinergic neurons might be particularly vulnerable to autoimmune attack in this population.

The latest work extends this line of evidence by looking at specific peptide sequences from alpha-synuclein, which is a protein famously noted to be aggregated in the neurons of Parkinson’s patients. It turns out that T-cells from such patients (and not from controls) recognize these peptides, and this process appears to drive a cytotoxic immune response. This would tie in very well with the MHC genetic connections, and may well be putting us towards a better, more comprehensive explanation of the whole disease.

In general, for diseases that seem to have both a genetic susceptibility component and an environmental exposure/history one, you’d have to think that there could well be an immune system mechanism involved. That’s the part of our bodies that most clearly responds to our own environmental exposures (thus the possibility of vaccination), and is (at the same time) genetically unique to each individual. Add in the way that immune system is capable of inflicting major continuing damage to whatever cell population it targets, and you have the scope to explain almost anything. But without hard data, that explanation isn’t worth much – just saying “Must be some autoimmune thing” doesn’t advance the field. Now, with Parkinson’s the hard data may well be coming in.

32 comments on “Parkinson’s As An Autoimmune Disease: More Evidence”

  1. Mike says:

    Has anyone done a study to see if patients on long-term immunosupressive therapy have a lower-than-expected incidence of Parkinson’s disease?

    1. Derek Lowe says:

      That’s a good thought. I believe that people have combed the data for other disease in this manner, but I can’t find anything in PubMed (at least not right off) that shows any such study in Parkinson’s. There are plenty of papers referencing immunosuppression, but always in the context of cell transplantation therapies.

      1. claire says:

        lower parkinsons risk in long term use of NSAIDs

        1. A3S says:

          I don’t think your reference supports that claim. They fail to find such a connection.

        2. tangent says:

          Huh? Their results superficially found an /increased/ risk, and they’re talking about whether it’s real.

    2. Immunobiologist says:

      Great idea. That was a major tip off for the role of the immune system in cancer.

  2. joy says:

    What’s the current thinking on a mechanism for reversing autoimmune diseases? Is anyone making credible progress towards finding a switch or a way to annihilate the specific responsible cell populations?
    I’d dig in the literature myself if I still had access (blissfully retired).

    1. GutDecipher says:

      There’s much work focusing on the interplay between immune cell repertoire and microbiome. Some people even have (anecdotally) found relief from symptoms of colitis/IBD/MS by intentionally infecting themselves with parasites; however it’s unclear if disease states can be completely sent into remission and how general or dangerous those treatments could be. Moises Velasquez-Manoff wrote a great, well-sourced book (An Epidemic of Absence) a few years ago focusing on many of the investigators and patients.

    2. TroyBoy says:

      Don’t forget about rituximab and its successor B-cell depleting antibodies. These antibodies do pretty good for certain forms of MS. So the question is: would B-cell depleting therapies work for Parkinsons?

    3. Mediocrates says:

      There’s currently a bit of a gold rush mentality surrounding immunometabolism, prompted by the observation that many activated immune cell types are very highly metabolic (up there with cancers), run specific metabolic programs (i.e. itaconic acid production in macrophages), and that there seems to be a rough divide between the metabolism of pro- and anti-inflammatory cells (aerobic glycolysis and fatty acid oxidation/OXPHOS, respectively.

      Since metabolic enzymes are classically among the most druggable targets, there’s a lot of potential for selectively inactivating certain inflammatory cells and/or promoting suppressor cells. There’s also a lot of interest in, uh, immuno-metabolo-oncology, since the tumor microenvironment is metabolically stressful to infiltrating lymphocytes and tends towards their inactivation, and therapies to restore immune cell metabolism might synergize with PD(L)-1-targeted therapies and the like.

  3. luysii says:

    Be careful to distinguish the inflammatory clearance of a dying cell from the problem causing it to die. Back in the day before people were sophisticated about the immune system, all we had to search for autoimmune disease were the classic immunoglobulin antibodies. People started finding them in a variety of autoimmune diseases, and for a while they were thought to be causative, until someone found them in strokes due to large intracerebral artery occlusions.

    Also autoimmune diseases tend to associate with each other, dermatomyositis with lupus etc. etc. I can’t recall a Parkinsonian patient of mine who had one, and I’ve probably treated more than 200 over the years.

    1. luysii says:

      Sorry I meant people began finding them in all sorts of neurologic diseases rather than all sorts of autoimmune diseases

  4. Barry says:

    Add Parkinson’s Disease to the list of possible adverse side-effects from “checkpoint” immuno-oncology interventions blocking CTLA4 or PD-1 or PD-1L.
    That may still be acceptable for a patient dying of cancer, but it argues against ever moving these therapies to the frontline.

  5. JB says:

    One interesting observation other groups has found is that post-translational modifications like glycosylation of IgGs are a unique marker for aging. Several IgG glycans (including FA2B, FA2G2, and FA2BG2) changed considerably with age and the combination of these three glycans can explain up to 58% of variance in chronological age, significantly more than other markers of biological age like telomere lengths. The remaining variance in these glycans strongly correlates with physiological parameters associated with biological age. IgG glycosylation appears to be closely linked with both chronological and biological ages. Considering the important role of IgG glycans in inflammation, and because the observed changes with age promote inflammation, changes in IgG glycosylation also seem to represent a factor contributing to aging and could contribute to inflammatory-Parksinson’s phenomena.

    1. Chris says:

      That’s fascinating- I’ve seen work about the connections between inflammation and aging, but not about IgG glycans. Is there any lab/paper you’d suggest to read up on it?

      1. JB says:


        Oncotarget. 2016 Feb 16;7(7):7455-68. doi: 10.18632/oncotarget.7059.
        Identification of novel plasma glycosylation-associated markers of aging.

        J. Proteome Res., 2016, 15 (6), pp 1853–1861
        Changes in Healthy Human IgG Fc-Glycosylation after Birth and during Early Childhood
        Noortje de Haan†, Karli R. Reiding†, Gertjan Driessen‡, Mirjam van der Burg§, and Manfred Wuhrer*†

        Medicine (Baltimore). 2016 Jul; 95(28): e4112.
        Profiling IgG N-glycans as potential biomarker of chronological and biological ages
        A community-based study in a Han Chinese population

        Glycoconjugate Journal 15, 683—689 (1998)
        Structural changes in the oligosaccharide moiety of
        human IgG with aging

        PLoS One. 2015 Nov 11;10(11):e0142645.
        Change in N-Glycosylation of Plasma Proteins in Japanese Semisupercentenarians

        And on the glycome-inflammation connection:

        Sci Rep. 2014; 4: 4347.
        Changes in IgG and total plasma protein glycomes in acute systemic inflammation

        Arthritis Rheum. 1988 Mar;31(3):333-8.
        Abnormal glycosylation of serum IgG from patients with chronic inflammatory diseases.

        Journal of Autoimmunity Volume 57, February 2015, Pages 1–13

        Glycans in the immune system and The Altered Glycan Theory of Autoimmunity: A critical review

        Current Opinion in Organ Transplantation: February 2011 – Volume 16 – Issue 1 – p 7–14
        The role of differential IgG glycosylation in the interaction of antibodies with FcγRs in vivo

  6. Peter S. Shenkin says:

    Parkinson’s is diagnosed based on symptoms by a process of elimination. Are we really sure all the patients are suffering from the same disease, under the surface? I don’t know how often a necroscopy is performed after death and how often that necroscopy exhibits the classical symptoms of PD (Lewy bodies, etc.). If the answer is “substantially less than all the time”, we ought to consider that what gets diagnosed as Parkinson’s may be any of several diseases. In some cases there is a known proximal cause (like infectious disease in Oliver Sacks’s “Awakenings” patients, or a drug overdose); in others, nothing like that can be identified. A distinction is sometimes made between Parkinsonism and Parkinson’s disease, but if the anatomical necroscopy always “looks the same” maybe it is all one disease. Then maybe, as Derek implies, these could all be different routes to a leaky blood-brain barrier, which then results in the autoimmune response. Or at least different routes to some other common mechanism.

    1. dstar says:

      “Parkinson’s is diagnosed based on symptoms by a process of elimination. Are we really sure all the patients are suffering from the same disease, under the surface?”

      This is a very, very important question, in my opinion.

      My late wife (diagnosed with throat cancer at 44 last august, passed away this March; I have to wonder if I’m the _only_ person to lose someone to cancer who doesn’t have irrational hatred towards ‘big pharma’ _or_ insurance companies) had fibromyalgia.

      Except I have a suspicion that fibromyalgia isn’t a disease, it’s a collection of symptoms with multiple underlying causes; there’s simply too much varience between what works for fibro patients even after you apply an incredibly pessimistic filter to eliminate the placebo effect.

      I would not at all be surprised to find that _any_ diagnosis of exclusion is, in fact, aggregating multiple underlying pathologies.

  7. Ron says:

    Thank you for this link to this new work. Very interesting.

    However, I think your MS (multiple sclerosis) link really does a disservice to the body of work that links MS (and SLE – Lupus) to Eptstein-Barr Virus. The epidemiology between those diseases and that virus is simply too hard to ignore despite the fact that it still is! Even after decades of research that confirms the association over and over and over.

    The only reason it is ignored is that the mechanism for how chronic reactivation of a mostly ubiquitous herpes virus in MS patients leads to autoimmune demyelination in susceptible patients. That is still speculative. Yes, there is a genetic link. But study after study has shown that the concordance of autoimmune diseases in identical twins is ~20-30% depending on the disease. Obviously, genetics play a role but not an (immuno)dominant one. Riddle me that?

    However, the link between EBV and MS and Lupus is much stronger than the link between cigarette smoking and lung cancer. Look at the odds ratios. The NMSS and other funding agencies continue to fund research into other areas that have already been largely disproven simply because the recipients of those funds are vested in the funding and who gets it.

    Anyone who peruses the work of Ascherio at Harvard (and many others) relating EBV and MS and chooses to ignore that association is, quite frankly, just putting their head in the sand. Yes, ~90% of the western world carries the virus. But 100.000% of the MS and Lupus population carry that virus. And when you are only screening 10 patients, the difference between 9/10 versus 10/10 doesn’t mean a lot. But when you screen 1000 patients (or much more), the difference between 900/1000 versus 1000/1000 means a whole lot. Look at the childhood Lupus data where the co-concordance is only about 60% versus 100%. You don’t have to be an epidemiologist to figure that out.

    Would a lytic inhibitor of EBV serve these populations? Maybe, based on the antigen expression that it might supress. Would a childhood vaccine against this virus wipe out MS and Lupus? I’ll bet my retirement on that one. Because if you don’t carry that virus, you have been misdiagnosed with either disease.

    Although I no longer work in medicinal chemistry, I get tired of hearing of how little we know about good drug targets when the drug discovery gurus ignore the ones that are staring them in the face when a first-year grad student in any field could read the data and ask, “Ummm, why hasn’t anyone worked on this?” This is the part of the field that I really, really don’t miss.

    And since this is the most widely read blog in the field, I would hope that we all get it right.

    1. TroyBoy says:

      Ron, people have been working a long time to make a vaccine for EBV with some progress, but not much success. I think it will difficult to make a vaccine that provides sterilizing immunity. So don’t bet your retirement yet.

      The problem with lytic inhibitors is that they’re only (semi-)successful when the virus is undergoing lytic replication. It would be better to go after a target critical for maintaining latent infection. Stay tuned…

      1. Ron says:

        Actually, SKB had “some” success with their EBV vaccine. I am not naive as to the hurdles for any vaccine directed against a herpetic virus. I would certainly bet my retirement when one exists and only then. But the fact that no work takes place in the area speaks volumes as to the magnitude of medical need, especially with the autoimmune implications where the epidemiology is so strong and also as to where resources are directed in drug discovery. Let’s not forget Burkitt’s lymphoma and PTLD with EBV either. My retirement is really not at risk.

        Agreed as to targets for maintaining latency. But, at least when I was in the field, that was still sort of a pipe dream. I’ll stay tuned…….but won’t hold my breath.

        In reality, and beyond the scope of this blog, there is very compelling evidence as to the autoantigen induced by EBV in MS. It just takes someone who is willing to (critically) evaluate the literature. Immunotolerance is a very valid approach for diseases where the autoantigen is known. That is admittedly a rarity. Why has there been little to no progress in Myasthenia gravis is beyond me. But once again, the whining about good drug targets still holds. Tell me how much drug discovery is being directed against the acetylcholine receptor in MG.

        I’ll stand by my assertions.

    2. tangent says:

      Interesting! I knew there was an association, but I had never heard it stated that the incidence in EBV-negative people is vanishingly small. I’ve seen ratios for sero+/sero- that are very substantial, like double-digit numbers, but you’re saying it’s actually “divide by zero”?

      Ascherio 2000 [pubmed/11021623] is referred to for this by several other articles, I don’t have the full text, but that abstract says odds ratio of 13.5 (95% CI = 6.3-31.4). Obviously there will be uncertainty that keeps the low end of the range finite, but why doesn’t the top end of it blow up?

      (Is there something about how these statistical calculations are done? Like if the observed odds are 10:10 and 0:20, I would naively expect the odds ratio’s central estimate and upper bound to be infinite. No?)

    3. dstar says:

      Remember: Correlation is not causation.

      It’s entirely possibly that whatever causes MS results in a susceptibility to EBV such that you are effectively guaranteed to be EVB positive.

      How likely that is? I don’t know.

      I _do_ know that if EBV doesn’t cause MS, and the correlation is truly there, investigating why you’re more likely to contract EBV if you’re predisposed to MS would likely be productive.

  8. steve says:

    So there’s some confusion here. What the quoted paper showed is that, if you take T cells from patients with Parkinson’s, they’ll respond to some peptides derived from a-synculin. This does not mean that the T cells are causative; they could have been generated secondarily to the initial insult. As noted in the paper, “Alleles of over twenty genes are associated with familial Parkinson’s disease, many of which encode proteins implicated in lysosomal degradation pathways including mitochondrial turnover.” So the jury is still out on what the initiating even is. With regard to EBV, it’s clear that the majority of people infected with EBV do not develop Parkinson’s. If it is involved it again does not mean that it’s causative as it could be that the immune response to EBV cross-reacts with dopaminergic neurons in some people causing an autoimmune reaction leading to Parkinson’s. Lots to sift out including some recent evidence that Parkinson’s actually begins with defects in folding of alpha-synculin in the gut that then moves up through the vagus as a prion. Vagotomy reduces the chance of Parkinson’s by 40%. The role of the microbiome in inducing poor protein folding in the gut is also being investigated. It’s clearly not a simple story.

  9. Ron says:

    Steve – Just an FYI: I never suggested a correlation between Parkinson’s and EBV……….I think that link would be very tenuous at best. And the reference to cross-reactivity between EBV and alpha-synuclein in the Wikipedia link above would be rather speculative at this juncture.

    I was speaking as to the links with MS and Lupus and to the hyperlink in this post for MS that ignores the overwhelming evidence already in the literature with regard to the environmental component (EBV) already linked to that autoimmune disease (as well as SLE) which has never been refuted in the scientific literature and is systematically ignored despite extensive reports supporting the association. The environmental component in the link in this post is not persuasive. In fact, the authors refer to their work in the abstract as a “hypothetical paper”.

    So if someone were to provide a legitimate link for MS and an environmental (viral) association, link to one that has endured extensive scientific scrutiny and not one where the authors describe their work as “hypothetical”.

  10. milkshaken says:

    I also think the inflammatory reaction and formation of antibodies to alpha-synuclein may well be a consequence of the pathology – not the cause. Dopamine-producing neurons are uniquely sensitive to oxidative damage. You inject primate with manganese salts, MPTP or rotenone, and the dopamine-producing neurons die out quite selectively (and you got Parkinson animal model). Sure there are transport mechanisms responsible for selectivity but I still think that crude mitochondrial poisons like these, resulting in elevated oxidative stress, and having uniquely parkinson-like effect, are telling us something important about the pathology.

    Could it be, for example, that some chaperone is not doing its job, misfolding of alpha-synuclein then increases oxidative stress which feeds back into alpha synuclein related pathology, and the reinforcement eventually produces Levy bodies, and glial cells start responding to the slowly dying dopamine-producing neurons, at which point there are also antibodies against alpha synuclein?

    1. tangent says:

      Does anyone have some idea why they’re like this? It seems like it must fall out of their metabolism, but does the dopamine synthesis pathway itself involve this hole to oxidative stressors? Something else about what these neurons do?

    2. Thoryke says:

      Thousands of gardeners have used Rotenone on their plants over the decades… Do clinical trial intake surveys or patient histories include that as a risk factor, I wonder?

  11. 2017 a nuclear heart scan has a chemical that is believed to cross the blood brain barrier for chemo intro. It produced a killer headache for me in 10/17 . When it was stopped/reversed, my headache subsided. I have Graves. My mother had dx for Parkinson’s. I’m 70, so probably not in my lifetime, but good luck to you guys/

Comments are closed.