1040-B Identifying tumor-associated antigens in different stages of multiple myeloma using low input HLA immunopeptidomics

BackgroundMultiple myeloma (MM) is a rare cancer that leads to malignant plasma cells. There are efforts to uncover how MM develops from precursor disease states, such as monoclonal gammopathy of undetermined significance (MGUS) and smoldering multiple myeloma (SMM). Characterizing the evolution of MM allows for earlier prognosis and intervention. Limited availability of tumor (CD138+) cells from patients poses challenges in studying MM stages. Immunopeptidomics typically requires a minimum of 50–100 million cells, exceeding what can typically be obtained from blood cancers. We optimized our HLA immunopurification protocol and paired it with ultra-sensitive mass spectrometry to enable analysis of sample inputs as low as 1 million cells. With this sensitive low input workflow, we characterized the immunopeptidomes of n=35 MM patients, across different stages of the disease.MethodsWe optimized our low input HLA immunopurification protocol by reducing lysis buffer volume, antibody concentration, protein G bead volume, bead wash volume, and desalting sorbent.1 2 Using HLA-I/II-expressing A375 cells, we refined the protocol with a range of cell inputs (0.5, 1, 2, 5, and 10 million cells). To confirm the method utility, we profiled a cohort of CD138+ MM samples (≤2 million cells) from a range of disease stages (MGUS; n=7, SMM; n=9, newly diagnosed multiple myeloma: NDMM; n=10, and relapsed/refractory multiple myeloma: RDMM; n=9). We used the Bruker trapped ion mobility spectrometry time of flight (timsTOF) single cell proteomics (SCP) instrument for low input HLA-I/II immunopeptidomics, interpreting spectra with Spectrum Mill against a database containing a reference human proteome and ribosome-sequencing-derived novel or unannotated open reading frames (nuORFs).3 ResultsOptimized low input immunopurification and ultra-sensitive mass spectrometry achieved ~700 HLA-I and ~650 HLA-II peptides on average across the cohort. In some samples, we identified up to ~4000 HLA-I and ~2000 HLA-II peptides. Given the depth, we found HLA-I nuORF-derived peptides, TAAs (e.g., BCMA, MMSET), and CTAs (e.g., MAGE family).ConclusionsHighly sensitive, low input immunopeptidome analysis enables the discovery of HLA peptides presented by primary MM tumor cells. We detected nuORF-derived peptides in the HLA-I and HLA-II immunopeptidome from MM samples for the first time. This optimized workflow and the resulting data may yield disease state biomarkers and immunotherapeutic targets.Ref