Poster Presentation HUPO 2019 - 18th Human Proteome Organization World Congress

Simple and efficient HLAp-ligandomics by using high-field asymmetric waveform ion mobility spectrometry (FAIMS) (#724)

Yuriko MINEGISHI 1 , Kensaku Nemoto 1 , Kazuma Kiyotani 1 , Koji Ueda 1
  1. Japanese Foundation for Cancer Research (JFCR), Koto-ku, TOKYO, Japan

 Human leukocyte antigen (HLA) class 1-binding peptide (HLAp) with somatic mutation, recognized as a neoantigen, is as an optimal target for cancer vaccine treatment. Though identifying actually-presented neoantigens is crucial for realization of cancer precision medicine, comprehensive analysis of neoantigens (HLAp-ligandomics) remains difficult due to insufficient HLAp isolation yield, undefined mass spectrometric parameters for non-tryptic digest peptides, and lack of bioinformatics pipelines to automatically construct “individual” proteome database from genetic sequence dataset.

 To overcome these difficulties, we optimized immuno-capturing procedures for HLAp isolation, employed high-field asymmetric waveform ion mobility spectrometry (FAIMS) for maximization of HLAp identification efficiency, and finally established a pipeline to construct proteome databases seamlessly from whole exome sequencing analysis. Particularly, FAIMS technology enables unique ion fractionation in vapor phase based on ion mobilities under different electric fields, allowing more sensitive and time-saving peptide identification in a single LC/MS run.

 The data was acquired with Orbitrap Fusion Lumos mass spectrometer installed with or without FAIMS Pro (Thermo Fisher Scientific). The compensation voltages (CVs) were set to -40/-60/-80 for targeting charge states 2 & 3, while they were set to -10/-20/-30 for targeting charge state 1. The HCT116-specific mutant proteome (mutanome) database was established by the in-house pipeline. As a result, the number of identified HLAp from 1.0E+7 HCT-116 cells (1,443 peptide groups) surpassed around twice as much as that without FAIMS (702 peptide groups). Majority of the identified HLAp (86.2%) had 8-10 amino acid length. Gibbs Cluster 2.0 revealed the dominance of A*01:01 and B*45:01-assigned peptides in HCT116. Importantly, a couple of neoantigens were successfully identified from the database search analysis using HCT116-specific mutanome.

 In the future, if high-performance HLAp profiling technology could be used in routine clinical practice, application of personalized cancer immunotherapy would be further accelerated.