Say, you have selected a number of linear B-epitopes from a protein sequence (or from several proteins) using your own selection criteria, and need to decide which one you should use for immunization (likely to result in a strong hyperimmune serum).
We will illustrate how to use B-scanner to analyse the epitopes in it by using data from a published article of Dakappagari et al, Cancer Res 2000;60:3782-3789. The ability of epitopes described in this article "to elicit weak or strong immune response" has been tested experimentally and shall allow us to control the quality of the in silico analysis.
Dakappagari et al wanted to find strong linear epitopes within the sequence of human ErbB2 to further incorporate them into their erbB2-vaccine. As they stated, the selection of “candidate B-cell epitopes expressed within the human HER-2 ECD was accomplished by computer-aided analysis using various correlates of protein antigenicity as reviewed by Kaumaya et al*”, and the exact criteria for selection were not disclosed.
_____
*Kaumaya, P. T. P., Kobs-Conrad, S., DiGeorge, A. M., and Stevens, V. De novo engineering of protein immunogenic and antigenic determinants. In: G. M. Anantharamaiah and C. Basava (Eds.), Peptides, Design, Synthesis & Biological Activity. pp. 133–164. Boston: Birkhauser, 1994
Thus, by certain methods, four epitopes were selected out of the human erbB2 sequence. Can we predict which one will be immunogenic and suitable for the vaccine or which epitope to raise in an erbB2-reactive hyperimmune serum?
Here we show how to choose the best epitope out of a group of peptide antigens using B-Scanner.
Since linear epitopes are generally 8-10a long and analysed sequences are at least twice as long, they may contain at least two independent epitopes. Hence, one approach will be to compare the best epitopes that can be found in the analysed sequences. Another approach being to compare the overall antigenicity of the entire sequences.
First, let's compare the overall antigenicity of the sequences (the comments on the settings are below):
Here we show the results as they were produced in Table format by B-scanner. The first analysis (a) was performed with Peptide Size 20 (the length of the shortest of the peptides), the results in (b) obtained with Peptide Size 12 which is the maximal size of a standard linear B-epitope including the flanking sequences. In both runs the results were requested as AGI BEST, which means that for every analysed sequence the best peptides of the defined Peptide Size will be shown, and the results will be sorted according to their AGI (antigenicity index)
To analyse the correctness of our prediction, lets compare the in silico rating of the peptide with the actual experimental data. Each of 4 peptides was used, in a carrier-coupled form, to immunize 2 rabbits (data from Dakappagari et al, Cancer Res 2000;60:3782-3789). Below we show the original data provided by authors, on the right – averaged response in 2 rabbits.
So. the predicted and actual epitope "strength" were:
Prediction: HER4 > HER1 >HER2 > HER 3
Experiment: HER4 > HER1 >HER2 > HER 3
As shown here, the predicted antigenicity for both 20 mers and 12 mers fully correlates with ability of the peptide to elicit specific antibodies. For 20-mers the prediction is more correct and correlates with P-value of <0.01 for all 4 peptides .
EpiQuest Suite and site www.epiquest.co.uk belongs to Aptum Biologics Ltd.
EpiQuest® is a registered Trademark of Aptum Biologics Ltd.
© 2018, 2020, 2021 Aptum Biologics Ltd.
