Information Technology Reference
In-Depth Information
Immunogenicity profiles of Human Proteins (Scored for the 8 most common human class
II
alleles)
Fig. 2.
Inherent immunogenicity of human proteins. To perform this analysis, we used the
EpiMatrix tool, which is a matrix-based algorithm for T-cell epitope mapping. This tool is
standardized so that comparisons can be made across predictions for different HLA alleles. In
this case we used the EpiMatrix prediction matrices for eight class II alleles that are represen-
tative of more than 98% of human populations. First, we measured the number of potential T
helper epitopes that would occur in random-sequence pseudo proteins composed of amino
acids at their naturally occurring frequencies and computed the mean “epitope score” per 1000
assessments of 0.5, with a standard deviation of ±7.9. We then compared proteins by summing
the total number of EpiMatrix scores for each frame (9 amino acids) that was above an ac-
cepted cutoff for immunogenicity (>1.67), and measured the difference between the scores for
“random” proteins and test proteins.
This type of immunogenicity analysis can be used to evaluate and compare pro-
tein therapeutics - to other protein therapeutics and to known antigens. Potential T-
cell epitopes are not randomly distributed throughout protein sequences but instead
tend to be clustered. These regions of unusually high binding potential are called “T-
cell epitope clusters,” “promiscuous epitopes,” or just “clusters” for short (Panina-
Bordignon, Tan, Termijtelen, Demotz, Corradin, and Lanzavecchia 1989). Epitope
clusters range from 9 to roughly 25 amino acids in length and, considering their
affinity to multiple alleles and across multiple frames, can contain anywhere from 4
to 40 binding motifs. ClustiMer, which is an optional feature of EpiMatrix, measures
and stores the MHC binding potential for 9 or 10 amino acid sequences to a number
of human HLA and then searches for extended regions containing high concentra-
tions of these high-scoring 9-mers.
A protein which scores “average” or even below average on our overall immuno-
genicity scale may still be immunogenic if it contains one or more clustered regions.
For example the well-known antigen tetanus toxin scores just +4 on the overall scale
but also contains several local regions with very high potentials (Fig. 3).
