Frontiers in Immunology (Apr 2018)

Discrimination Between Human Leukocyte Antigen Class I-Bound and Co-Purified HIV-Derived Peptides in Immunopeptidomics Workflows

  • Thomas Partridge,
  • Annalisa Nicastri,
  • Anna E. Kliszczak,
  • Louis-Marie Yindom,
  • Benedikt M. Kessler,
  • Nicola Ternette,
  • Nicola Ternette,
  • Persephone Borrow

DOI
https://doi.org/10.3389/fimmu.2018.00912
Journal volume & issue
Vol. 9

Abstract

Read online

Elucidation of novel peptides presented by human leukocyte antigen (HLA) class I alleles by immunopeptidomics constitutes a powerful approach that can inform the rational design of CD8+ T cell inducing vaccines to control infection with pathogens such as human immunodeficiency virus type 1 (HIV-1) or to combat tumors. Recent advances in the sensitivity of liquid chromatography tandem mass spectrometry instrumentation have facilitated the discovery of thousands of natural HLA-restricted peptides in a single measurement. However, the extent of contamination of class I-bound peptides identified using HLA immunoprecipitation (IP)-based immunopeptidomics approaches with peptides from other sources has not previously been evaluated in depth. Here, we investigated the specificity of the IP-based immunopeptidomics methodology using HLA class I- or II-deficient cell lines and membrane protein-specific antibody IPs. We demonstrate that the 721.221 B lymphoblastoid cell line, widely regarded to be HLA class Ia-deficient, actually expresses and presents peptides on HLA-C*01:02. Using this cell line and the C8166 (HLA class I- and II-expressing) cell line, we show that some HLA class II-bound peptides were co-purified non-specifically during HLA class I and membrane protein IPs. Furthermore, IPs of “irrelevant” membrane proteins from HIV-1-infected HLA class I- and/or II-expressing cells revealed that unusually long HIV-1-derived peptides previously reported by us and other immunopeptidomics studies as potentially novel CD8+ T cell epitopes were non-specifically co-isolated, and so constitute a source of contamination in HLA class I IPs. For example, a 16-mer (FLGKIWPSYKGRPGNF), which was detected in all samples studied represents the full p1 segment of the abundant intracellular or virion-associated proteolytically-processed HIV-1 Gag protein. This result is of importance, as these long co-purified HIV-1 Gag peptides may not elicit CD8+ T cell responses when incorporated into candidate vaccines. These results have wider implications for HLA epitope discovery from abundant or membrane-associated antigens by immunopeptidomics in the context of infectious diseases, cancer, and autoimmunity.

Keywords