Painting of TcR interacting with artrificial membranes by Raghuveer Parthasarathy
TcR interacting with artificial membrane1

Why does autoimmune disease (sometimes) follow viral infection?2

It’s a pretty well-known phenomenon, but a definite answer isn’t yet known — and of course there may not be a single answer, there may be multiple causes. We know that many autoimmune diseases seem to be triggered by some sort of infection or inflammation. A classic example is Guillan-Barre syndrome, which is a little more common (though still very rare) in people who have received certain influenza vaccines, but there are plenty of other examples.3 It’s not believed that the infection actually “causes” the disease, but rather that someone who already has a genetic predisposition to the autoimmune disease needs to have some kind of environmental trigger to have the disease actually kick in; and, very rarely, a viral or other infection will provide that trigger.

(The genetic predisposition is clear because, among other points, identical twins are much more likely to both get autoimmune disease than are fraternal twins; whereas the need for an environmental trigger is clear because even if your identical twin gets an autoimmune disease, you’re usually less than 50% likely to get it yourself. Note that I’m lumping together hundreds of different diseases into the “autoimmune” package, and the specific odds and so on differ for each one.)

OK, so if you have a genetic predisposition to autoimmunity — and let’s get more specific, the paper I’m looking at deals with multiple sclerosis (MS) — there’s a small chance that a viral infection will trigger that disease. One of the most popular models for this is “molecular mimicry”. Simplified: This is the notion that a viral protein looks, to a T cell, a little bit like a self protein. The viral protein appears in the context of infection, with its concomitant inflammation and tissue damage and so on, and the T cell is activated to it. The T cell wouldn’t be activated by the self protein because it hasn’t been seen in the context of inflammation before, but once over the activation hurdle the T cell is now able to attack the self protein, and this is autoimmunity.

TcR/MHC
T cell receptor (top) interacting with MHC

Molecular mimicry is an attractive model, but there’s not a lot of direct evidence for it.  Another possibility has been proposed for a while: Dual TcRs. Normally, T cells can only recognize a single target. This is by “design”;4 if the T cell can see two targets, it could get activated by one, and then attack the other, even if the second target was never present during inflammation. This sort of dual target recognition is obviously dangerous, and there are safeguards that mostly prevent it; but some T cells do sneak through with at least the theoretical potential for dual recognition. So what could happen here is that one TcR could be directed against the pathogen, and activate the T cell; then the other TcR, recognizing self, could run amok because it’s now on an activated T cell.

T cells with dual specificity do exist, at a fairly significant frequency (1-8%; at least one source claims as high as 33%, which seems much too high to me), but whether they actually do anything in autoimmunity is up in the air. This idea has been around for a while, but I don’t think there’s been much evidence for it happening naturally. In at least one case, where it was tested in an artificial system, dual TcRs did not seem to be responsible for an automimmune disease. 5

The most recent paper offers evidence that (in quite an artificial system) dual-specificity T cells are responsible for multiple sclerosis: 6

Our results demonstrate the importance of dual TCR–expressing T cells in autoimmunity and suggest a mechanism by which a ubiquitous viral infection could trigger autoimmunity in a subset of infected people, as suggested by the etiology of multiple sclerosis.

It’s an interesting and solid paper as far as it goes, but we’re left with the issue of this being a highly artificial system — mice with manipulated TcRs and manipulated autoimmune disease. Is this a real issue in natural autoimmunity and natural infections? This paper doesn’t really address that, but it does support the notion that it’s something to look more closely at.  (And again, different autoimmune diseases, or even different people with the same disease, may have altogether different triggers.  Maybe some people have molecular mimicry as the trigger while others have dual TcRs and other have who knows what.)


  1. By Raghuveer Parthasarath, then in the Groves lab[]
  2. Also, why are so many of my keyboard keys sticking together? An altogether easier question quickly answered by pointing to my kids “helping” me with my work while holding popsicles[]
  3. For a review:
    Fujinami, R. (2001). Can Virus Infections Trigger Autoimmune Disease? Journal of Autoimmunity, 16 (3), 229-234 DOI: 10.1006/jaut.2000.0484[]
  4. I.e. evolution.[]
  5. McGargill MA, Mayerova D, Stefanski HE, Koehn B, Parke EA, Jameson SC, Panoskaltsis-Mortari A, & Hogquist KA (2002). A spontaneous CD8 T cell-dependent autoimmune disease to an antigen expressed under the human keratin 14 promoter. Journal of immunology (Baltimore, Md. : 1950), 169 (4), 2141-7 PMID: 12165543 []
  6. Ji, Q., Perchellet, A., & Goverman, J. (2010). Viral infection triggers central nervous system autoimmunity via activation of CD8+ T cells expressing dual TCRs Nature Immunology, 11 (7), 628-634 DOI: 10.1038/ni.1888[]