EWS-FLI-1-targeted cytotoxic T-cell killing of multiple tumor types belonging to the Ewing sarcoma family of tumors

Abstract

Purpose: The Ewing sarcoma family of tumors (ESFT) comprises a group of aggressive, malignant bone, and soft tissue tumors that predominantly affect children and young adults. These tumors frequently share expression of the EWS-FLI-1 translocation, which is central to tumor survival but not present in healthy cells. In this study, we examined EWS-FLI-1 antigens for their capacity to induce immunity against a range of ESFT types.

Design: Computer prediction analysis of peptide binding, HLA-A2.1 stabilization assays, and induction of cytotoxic T-lymphocytes (CTL) in immunized HLA-A2.1 transgenic mice were used to assess the immunogenicity of native and modified peptides derived from the fusion region of EWS-FLI-1 type 1. CTL-killing of multiple ESFT family members in vitro, and control of established xenografts in vivo, was assessed. We also examined whether these peptides could induce human CTLs in vitro.

Results: EWS-FLI-1 type 1 peptides were unable to stabilize cell surface HLA-A2.1 and induced weak CTL activity against Ewing sarcoma cells. In contrast, peptides with modified anchor residues induced potent CTL killing of Ewing sarcoma cells presenting endogenous (native) peptides. The adoptive transfer of CTL specific for the modified peptide YLNPSVDSV resulted in enhanced survival of mice with established Ewing sarcoma xenografts. YLNPSVDSV-specific CTL displayed potent killing of multiple ESFT types in vitro: Ewing sarcoma, pPNET, Askin's Tumor, and Biphenotypic sarcoma. Stimulation of human peripheral blood mononuclear cells with YLNPSVDSV peptide resulted in potent CTL-killing.

Conclusions: These data show that YLNPSVDSV peptide is a promising antigen for ESFT immunotherapy and warrants further clinical development.

Figure 1. Binding, peptide-mediated MHC stabilization, and CTL induction of native and modified peptides

(A–C) Three native peptides (red bars) and 21 modified peptides (blue bars) were analyzed for binding to HLA-A2.1 using the REVEAL MHC-Peptide Binding Assay. Peptide sequences are shown in Table 2: (A) Native peptide SYGQQNPSY (#1) and its modified derivatives, (B) Native peptide SSYGQQNPS (#9) and its modified derivatives, (C) Native peptide QQNPSYDSV (#14) and its modified derivatives. Each peptide was given a score reported quantitatively as a percentage of the signal generated by the test peptide versus a positive control peptide (GILGFVFTL, Flu. open bars). (D–F) The ability of native peptides (D) SYGQQNPSY (#1), (E) SSYGQQNPS (#9), (F) QQNPSYDSV (#14) and their modified derivatives to stabilize HLA-A2.1 molecules. The Fluorescence Index is the median fluorescence intensity (MFI) with the indicated concentration of test peptide divided by the MFI in the absence of peptide. (G–I) CTL induction in HLA-A2/Kb transgenic mice by native peptides (G) SYGQQNPSY (#1), (H) SSYGQQNPS (#9), (I) QQNPSYDSV (#14) and their modified derivatives was assessed through the killing of TC-71 Ewing’s Sarcoma target cells. Graphs show a representative experiment from multiple independent experiments with similar results. Error bars represent SD.

Figure 2. Adoptive transfer of T-cells into mice with established xenograft tumors

SCID/beige mice were challenged with viable TC-71-Luc cells i.v. and treated on days 5 and 12 with CD8 T-cells from immunized HLA-A2/K^b mice (n=4). (A) Bioluminescence imaging of mice 21 days post tumor challenge with: (i) PBS alone (No T-cells), (ii) EWS-FLI-1 (YLNPSVDSV-specific) T-cells, and (iii) Flu (GILGFVFTL-specific) T-cells. (B) Mouse survival post tumor challenge. Treatments are represented by arrows. *indicates significant increase (p<0.05) in survival of mice recipient of EWS-FLI-1 (YLNPSVDSV-specific) T-cells over both control groups. (C) Tumor frequency in autopsied mice. (D) MRI imaging of abdominal tumor mass in a mouse recipient of EWS-FLI-1 (YLNPSVDSV-specific) T-cells.

Figure 3. Peptide-mediated CTL killing of multiple ESFT targets

(A) CTL were induced in HLA-A2/K^b mice by immunization with native peptide QQNPSYDSV (#14) or its modified derivative YLNPSVDSV (#24). The capacity of CTL to kill 11 EWS-FLI-1 Type 1+, HLA-A2.1+ ESFT targets (described in Table 3) was subsequently assessed. Graphs show combined data from 3 independent experiments with similar results. Error bars represent SEM. (B–C) Healthy human PBMC from 3 different HLA-A2.1+ donors were stimulated with native- QQNPSYDSV (#14) or modified - YLNPSVDSV (#24) peptide for 4 weeks to generate CTL. (B) Expansion of human CTL lines. Pentamer-positive staining as a percentage of total CD8 cells. Closed squares = YLNPSVDSV, open triangles = GILGFVFTL, B = Before culture, A = After culture. Graphs show combined data from 3 independent experiments. Error bars represent SEM. (C) EWS-FLI-1 Type 1-targeted cytotoxic activity using TC-71 cells as targets. White bars = QQNPSYDSV (#14), Black bars = YLNPSVDSV (#24). Graphs show combined data from 2 independent experiments. Error bars represent SD.