Premium
An Approach to Searching Protein Sequences for Superfamily Relationships or Chance Similarities Relevant to the Molecular Mimicry Hypothesis: Application to the Basic Proteins of Myelin
Author(s) -
Weise Michael J.,
Carnegie Patrick R.
Publication year - 1988
Publication title -
journal of neurochemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.75
H-Index - 229
eISSN - 1471-4159
pISSN - 0022-3042
DOI - 10.1111/j.1471-4159.1988.tb03096.x
Subject(s) - molecular mimicry , biology , myelin basic protein , mimicry , genetics , myelin , computational biology , protein sequencing , peptide sequence , immunology , neuroscience , antigen , gene , ecology , central nervous system
A rapid method for similarity searches (FASTP program) was used to identify similarities between a protein database and the human basic proteins from myelin [P 2 protein and 17.2K, 18.5K, and 21.5K variants of myelin basic protein (MBP)]. From similarity scores, we concluded that none of the presently known proteins are in a family containing the MBPs. No new members were found for the lipid‐binding family of which P 2 is a member. Sequence similarities deemed relevant to the molecular mimicry hypothesis for virus‐induced autoimmunity were identified in FASTP data with the aid of microcomputer programs. Several MBP/viral protein similarities were found that have not been reported previously. Of note because of their association with demyelinating conditions were proteins from visna and vaccinia. Similarity with visna was specific to the 21.5K and 20.2K MBPs. The most interesting new possibility for mimicry involving the P 2 protein was between the influenza A NS2 protein and a sequence region of P 2 thought to be neuritogenic in animals and mitogenic for lymphocytes from some patients with Guillain‐Barre syndrome (GBS). This may have relevance for some cases of GBS associated with the 1976 U.S.A. swine flu vaccination program. Because FASTP reports only the best similarities between proteins, searches with FASTP may not have detected all the examples of mimicry present in the database. Searches might also be more effective if similarities could be scored on immunological rather than structural relatedness.