Successful Generation of Cytotoxic T-Cells That Kill CLL Cells Using Heteroclitic Peptides Is Independent of the Native Peptide Binding Affinity to HLA-A*0201.

Targeting immunoglobulin (Ig) framework region (FR) derived peptides offers the advantage of a less patient specific immunotherapeutic strategy in B-cell malignancies. A major limitation of this method is the generally low immunogenicity and low binding affinity of these peptides to MHC class I and class II molecules. Heteroclitic peptide modifications can increase immunogenicity of low binding peptides while leaving T-cell recognition residues intact, and improve ability to generate cytotoxic T cells lines (CTL). It is not known whether such CTLs can still kill tumor cells that express native peptides below a lower threshold of binding affinity. To address this, we sequenced Ig, identified nonameric and decameric peptide sequences that potentially bind to HLA-A*0201 (HLA-A2) that were frequently shared among patients. We used two independent computer prediction analysis tools, determined binding using the T2-binding assay, screened peptide specific CTL responses with an established T-cell expansion system, and assessed cytotoxicity of the CTL lines against native and heteroclitic peptide pulsed APCs and also primary tumor cells from which the native peptides were derived. 34 FR-derived peptides were synthesized, 17 native peptides selected to represent a wide range of predicted binding to HLA-A2, and 17 corresponding heteroclitic counterparts. There was a strong correlation of predicted binding assessed by both scores (Spearmen rho=0.62; p=0.0001) and with respective T2-binding (Spearmen rho=0.66; p<0.0001). Heteroclitic peptides has significantly enhanced predicted binding compared to their native counterparts (Parker Score p<0.00001; Rammensee Score p=0.004, T2-binding p=0.0005) and CTLs generated against heteroclitic peptides had significantly enhanced killing of CD40 activated B-cells pulsed not only with the corresponding heteroclitic peptide (p=0.0003), but also with the native peptide (p=0.04). The binding affinity of low (n=10; FI<0.5) or intermediate binding (n=7; FI>0.6) native peptides did not correlate with specific lysis of peptide pulsed CD40 activated B-cells by CTLs generated against native peptides. Binding affinity of heteroclitic peptides correlated with the ability to induce CTL responses (Spearman rho=0.50; p=0.04). CTLs generated against heteroclitic peptide killed primary CLL cells more effectively than those generated against their native counterparts (p=0.01). Most importantly, the specific lysis of primary tumor cells by successfully generated CTLs was independent of the original binding affinity of the native peptides, both as measured by T2 binding assay (Spearman rho=0.38; p=0.36) and as predicted by the Parker Score (Spearman rho=0.22; p=0.60) or the Rammensee Score (Spearman rho=0.02; p=0.95). Thus, the present study demonstrates recognition of naturally processed Ig-derived peptides even with extremely low binding affinity to MHC class I when higher affinity analogues are used as ‘in vitro’ immunogens. Once CTLs are generated, cytotoxicity appears to be independent of the original binding affinity, suggesting that the rate limiting factor is the ability to generate the immune response, but that once generated, these CTLs have ability to kill tumor cells bearing even very weakly immunogenic peptides. These findings have significant implication for vaccination strategies in B-cell malignancies and warrant in vivo evaluation of this model in CLL.