Ana Marcu
Published: 2020
Total Pages: 0
Get eBook
Over the past two decades, immunotherapy has become the fourth pillar in cancer treatment alongside surgery, radiotherapy, and chemotherapy. Immune checkpoint blockade shows objective clinical success but only a fraction of patients benefit from tumor regression in response to immunotherapy. The number of immune-oncology-related clinical trials initiated each year is increasing. And yet, NY-ESO-1, is the most targeted antigen type (31 clinical trials), and melanoma the most targeted tumor type (33 clinical trials)1. Thus, studying underrepresented cancers and characterizing novel antigens is essential for the development of new treatment options. T cells are the main mediators of anti-tumor immune responses and they recognize human leukocyte antigen (HLA)-presented peptides. The entirety of HLA-presented peptides is termed the immunopeptidome. This work addresses three aspects of antigen discovery for T-cell based immune-oncology approaches. The Results Part I section presents the HLA Ligand Atlas, a collection of benign HLA-I and -II immunopeptidomes covering 21 human subjects and 29 tissue types. We observe that immunopeptidomes are more similar within individuals as opposed to the expected tissue-based clustering reported for transcriptomes2,3 and proteomes4. Furthermore, cryptic HLA ligands have been previously reported only in tumors and cell lines. In this dataset, we identified 1,407 cryptic peptides, showing that they are not per se tumor exclusive. Based on a case study in three glioblastoma patients, we illustrate the applicability of the multi-tissue HLA Ligand Atlas dataset to prioritize tumor-associated antigens for downstream experimental validation. The Results Part II section describes the theoretical requirements for the use of mass spectrometers (LC-MS/MS) in immunopeptidomics experiments. In addition, a benchmark based on a serial dilution of a JY peptide eluate spiked with heavy-labeled peptides across five different LC-MS/MS systems reports performance differences and complementarity to the in house Orbitrap Fusion Lumos. These results showed that the timsTOF Pro from Bruker Daltonics and the q Exactive HF from Thermo Fisher Scientific could be suitable for immunopeptidomics experiments. The results Part III section focuses on mapping the immunopeptidomic landscape of breast cancer, paving the way for generating a warehouse of shared tumor-associated antigens for frequent HLA allotypes. The selection of warehouse peptides will be the basis for a future phase I/II clinical study testing safety and toxicity of peptide vaccination. This project is still ongoing, as data acquisition has not been finalized. Data reanalysis and a stronger focus towards the triple negative breast cancer subtype will reshape the outcome of this project. Thus, this work addresses three major steps related to the discovery of T cell-based antigens for cancer immunotherapy: i) an in-depth consideration of the mass spectrometric instrumentation required for immunopeptidomics studies, ii) the generation of a benign multi-tissue draft of the human immunopeptidome, which is required for the definition of tumor-association, and iii) mapping the breast cancer immunopeptidome and the definition of tumor-associated antigens.