With the availability of complete genome sequences from many microbes, it is now possible to learn, purely through database analysis, which species carry a peptide sequence of interest.
Blank and colleagues previously found a six-amino acid antigenic determinant associated with antibodies from people with the autoimmune disorder antiphospholipid syndrome (APS).
Since close matches to this short sequence exist in proteins from a variety of common human pathogens, they set out to test the possibility that molecular mimicry between the bacterial proteins and the endogenous host protein explains the loss of self-tolerance in APS patients.
Blank et al. treated mice with extracts from several bacteria and a pathogenic yeast, most of which they knew to express such a sequence.
Several of the target sequence-expressing bacteria induced high-affinity antibodies specific for the same autoantigens identified in APS patients.
For at least two of the bacterial species, these antibodies were clearly pathogenic, in that, when purified, concentrated, and transferred to naïve recipient mice, they could cause symptoms of APS, particularly a prothrombotic phenotype and a high rate of spontaneous abortions.
Curiously, the mere presence of these antibodies was not sufficient to cause autoimmune disease, since the mice that produced these antibodies showed no such symptoms.
Along with their various bacterial preparations, Blank et al. challenged the animals with the tetanus toxoid protein, which contains three sequences that are similar in their 3-dimensional structure to the target peptide.
Because this treatment also induced pathogenic autoantibodies, it may be important to test whether this widely administered vaccine confers any risk of APS to humans.
Contact: John Ashkenas
Journal of Clinical Investigation