EpiQuest-B | EpiQuest-A | EpiQuest-IM | EpiQuest-C |EpiQuest-T | EpiQuest-H | EpiQuest-M | In Charge | EpiStat
Introduction
There are various approaches available for predicting T-epitopes, particularly CTL epitopes. Many of these approaches focus on predicting a peptide's ability to bind to the groove of MHC molecules, indicating its potential presentation in the context of MHC class I or II, respectively.
However, it's worth noting that less than 5% of peptides binding to MHC demonstrate any biological activity, and even fewer are strong, functional epitopes. Furthermore, when planning to use epitopes for immunization, it is highly desirable to utilize only immunodominant sequences. This is because an excess of weak epitopes can lead to the suppression of the immune response (see Chung, B., Stuge, T.B., Murad, J.P., Beilhack, G., Andersen, E., Armstrong, B.D., Weber, J.S., Lee, P.P., 2014. Antigen-Specific Inhibition of High-Avidity T Cell Target Lysis by Low-Avidity T Cells via Trogocytosis, Cell Reports 8, 871–882).
In contrast to other programs, EpiQuest-T predicts areas in the molecule that are likely to contain strong, immunodominant CTL epitopes. This focus on identifying immunodominant sequences can enhance the efficacy of epitope selection for immunization purposes.
Strong and Weak T-epitopes
Indeed, while much effort in informatics is currently directed towards defining peptide T-epitopes capable of binding to the groove of MHC class I or II molecules, it's crucial to consider the role of the T-cell receptor (TCR) as well. The TCR also determines the strength or affinity of binding to the peptide in the context of the MHC.
Drawing from our experience with sequences capable of eliciting high-affinity antibodies, we have shifted our focus to the structural characteristics of peptide epitopes. Specifically, we examine whether a peptide epitope is capable of eliciting a strongly binding TCR. This approach recognizes that T (CTL) epitope sequences can differ in their capability to induce strongly binding TCR, irrespective of their ability to bind well to the respective MHC.
Therefore, in the search for sequences for T-cell epitopes, the program should identify the unique features of the sequence that allow it to elicit TCRs that would strongly bind to it in the context of the respective MHC. This comprehensive approach ensures the identification of epitopes that are not only MHC-binding but also capable of eliciting strong TCR binding, enhancing their potential as immunogens.
The Scheme on the left demonstrates that the stronger the binding of the TCR to the peptide (in the context of MHC), the stronger the stimulation of the respective clone. The scheme is presented after: El Shikh, M.E.M., El Sayed, R.M., Sukumar, S., Szakal, A.K., Tew, J.G., 2010. Activation of B cells by antigens on follicular dendritic cells. Trends Immunol 31, 205–211. https://doi.org/10.1016/j.it.2010.03.002
The Algorithm
EpiQuest-T utilizes our unique algorithm, which has identified the typical structural features of a strong T-epitope recognized by the T-cell receptor (TCR) in the context of a particular type of MHC molecule.
The algorithm calculates the value of an individual amino acid based on its context using a matrix of values. This matrix determines the relative importance of each amino acid in the sequence by considering its close and distant neighbors. It was developed experimentally through the analysis of the context for over 10,000 amino acids in known epitopes, including those that are strong, weak, and non-immunogenic.
The software analyzes the linear protein sequence and calculates the probability of an amino acid at a given position in the sequence, depending on the context, to be involved into high affinity (immuno-dominant) CTL epitope depending on its context.
Program output
The program predicts areas likely to contain immunodominant CTL epitopes and ranks them according to their strength. The results are presented as a histogram and in tabular format. It's important to note that the prediction is haplotype-restricted, and the peptide is recognized by the T-cell receptor (TCR) in the context of a particular MHC, which contributes to the complex recognition process.
In the example presented below, which analyzes the presence of HLA-A2 restricted CTL epitopes in an insulin sequence, the program identifies several areas with potential epitopes of varying strength. It effectively discriminates strong epitopes from both weak and negative areas, with the main peak corresponding to the immunodominant CTL epitope in the insulin sequence. The known epitopes are indicated below the histogram for reference.
The tabular results show relative immunogenicity (AGI) and antigenicity corrected for the length of the fragment (AGR).
As can be seen, the immunodominant epitope-containing sequence has the highest predicted score.
EpiQuest-T and other programs
To illustrate the key differences between EpiQuest-T and programs based on finding the best MHC binder, we have compared the prediction results between EpiQuest-T and of the program suggested by IEDB in the optimal settings. In both examples shown below the red boxes indicate areas potentially containing functional CTL epitopes, blue boxes indicate the best (highest score) epitopes predicted by IEDB tool, and the purple ones - the actual biologically active epitopes discovered experimentally. Orange frame indicates the best epitope according to either the prediction or to experimental data.
As observed, the predictions made by EpiQuest-T directly cover the area of the most functionally active CTL epitope in the LCV glycoprotein. In contrast, among the predictions made by the IEDB tool (with the 14 best predictions shown), none correspond to actual functional CTL epitopes in this sequence.
While it's possible that some of the IEDB predictions may appear among the lower score predictions, this example clearly indicates a lack of connection between the predictions made by the program and the functional activity of the CTL epitopes. This underscores the importance of utilizing tools like EpiQuest-T that focus on identifying epitopes with strong functional activity.
For another protein, EpiQuest-T accurately defines an area that corresponds to the most likely location of the immunodominant epitope, covering exactly the locations of three functional CTL epitopes in this protein. In contrast, among the eight best predictions made by the IEDB tool, the lowest one corresponds to only a part of the area where the functional epitopes are located. Additionally, the location of the best epitopes predicted by the IEDB tool does not contain any functional epitopes at all.
This comparison highlights the superior accuracy of EpiQuest-T in identifying epitopes that align with functional CTL epitopes, emphasizing its effectiveness in epitope prediction.
The presented examples were analysed for mouse CTL epitopes, H2kB haplotype.
AGI index as predictor of CTL epitope "strength"
The program EpiQuest-T, along with its T-Scanner version, which is more adapted for the analysis of large collections of epitopes, assigns a value of potential "strength" to each epitope. This is done through the Antigenicity Index (AGI), which enables comparison of epitopes and selection of the best ones. Below, we provide two examples demonstrating how the AGI index is related to the functional activity of CTL epitopes.
In this example, we compared the functional activity of several mutant epitopes to their AGI values defined by EpiQuest-T. The mutant epitopes exhibited varying abilities to activate splenocytes, and we found that their functionality directly correlated with the AGI values determined by the program.
For this analysis, we utilized two different matrices offered by EpiQuest-T for the H2kB haplotype: Matrix H1, which exhibits a selective preference for high-activity epitopes, thereby discriminating them better from low and intermediate activity epitopes; and Matrix HI2, which demonstrates better recognition of high and intermediate activity epitopes. This comparison allowed us to assess the relationship between epitope activity and their AGI values across different matrix settings.
All H2kB epitopes from the IEDB database were categorized based on their activity in functional tests, as described in the database. The AGI index was calculated by EpiQuest-T for the best 8-mer epitope within each epitope sequence, which could range from 8 to 15 amino acids. The figure above displays the values for epitopes in each group and the median AGI value for every group.
The low values observed for the best epitopes of the H2kD haplotype indicate that the analysis is specific to this particular haplotype, highlighting the MHC-restricted nature of the data.
Analysis of Demo Sequences
For a full overview of the program's capabilities you can use Demo Mode. The Manual for EpiQuest-T explains how to modify settings. You can also see the examples of how to analyse the sequences and interpret the results by visiting the Demo Sequences analysis page.
Scanning CTL epitopes with T-Scanner
When you need to rank epitopes within a selected group, especially when dealing with a large number of epitopes (up to 2000 in one file), you can utilize T-Scanner.
T-Scanner is based on the same algorithm and matrix as EpiQuest-T. However, it is specifically developed for comparing a large number of epitopes. These epitopes can be selected by other programs, such as those predicting MHC-binding or proteolytic cleavage of the sequence in MHCI compartments, or they can be eluted from target cells, among other sources. T-Scanner facilitates efficient ranking and comparison of epitopes, enabling researchers to identify the most promising candidates for further study or application.
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.
