The use of reverse immunology to identify HLA-A2 binding epitopes in Tie-2

Abstract

A potential target for a cancer vaccine would be receptors, such as Tie-2 which are over expressed on tumour endothelium. Using computer aided motif predictions for possible HLA class I epitopes, we have identified peptides from Tie-2 that should bind with a range of affinities to HLA-A*0201. No direct correlation between predicted values and actual binding affinities was observed. Although, the programs did produce a number of false positives, two epitopes were predicted that bound with relatively high affinity when compared with an influenza peptide. We have previously identified a Tie-2 epitope and shown that it was only immunogenic when we substituted preferred amino acids at key anchor residues to increase binding affinity. In this study we used a similar approach to generate modified epitopes. When HLA-A2 transgenic mice were immunised with peptides, CTL killing of the target cells was only achieved when the wild type epitope was presented at moderate levels. Moreover, the efficiency of immunisation was increased when we linked CD4 epitopes to CD8 epitopes. Caution should therefore be employed in the use of both reverse immunology and anchor modification of CTL epitopes in the identification of CTL epitopes for cancer vaccines.

Fig. 1

No correlation between predicted binding scores and T2 stabilisation Computer algorithms (i) SYPEITHI or (ii) Parker were used to predicted possible epitopes. T2 stabilisation was determined by incubating T2 cells with 100 μg/ml peptide overnight at 37°C in serum free medium with β₂microglobulin. After a 1 h incubation with brefeldin A, cells were immunolabelled with an anti-HLA-A*0201 antibody. T2 stabilisation = [(MLF with peptide − MLF without Peptide)/MLF without peptide] + standard deviation. (MLF mean linear fluorescence) A score of 1 would indicate no increase in MHC stabilisation above background

Fig. 2

Comparison of T2 MHC stabilisation of Tie-2 modified peptides compared to native peptides. T2 cells were incubated with 100 μg/ml peptide overnight at 37°C in serum free medium with β₂microglobulin. After a 1 h incubation with brefeldin A, cells were immunolabelled with an anti-HLA-A*0201 antibody. Both native and modified peptides are shown on the y-axis against T2 binding score. T2 binding results are the [(MLF with peptide − MLF without Peptide)/MLF without peptide] + standard deviation. (MLF mean linear fluorescence) A score of 1 would indicate no increase in MHC stabilisation above background. This base line is indicated on the graph

Fig. 3

Immunisation of HLA-A2 mice with mZ9 peptide. HLA-A*0201 transgenic mice were vaccinated three times at fortnightly intervals with 100 μg Tie-2 mutant peptide (mZ9). Splenocytes were harvested and restimulated in vitro with the appropriate modified peptide for 6 days. CTL activity was assessed in a ⁵¹Cr release assay using T2 cells pulsed with either native (filled triangle) or modified peptide (filled inverted triangle) as targets. The data represents the average percentage specific lysis (n=3)