Mass spectrometry-based antigen discovery for cancer immunotherapy
Introduction
Cancer immunotherapy amplifies or reprograms the inherent capacity of immune system to recognize molecular entities expressed specifically on tumor cells and to eliminate the cells. Cancer specific antigens that are presented as human leukocyte antigens (HLA) binding peptides (HLAp) on the surface of cells, namely the immunopeptidome, serve as the leading targets and they may be derived from tumor-associated (over)expressed self-proteins, from oncogenic viruses, endogenous retroviral elements, or mutated tumor proteins. The remarkable clinical results of the immune checkpoint blocking therapies has motivated researchers to discover the immunogenic cancer-specific antigens that mediate T cell-based killing and long-lasting disease control [1]. Such antigens may be further exploited in the development of personalized vaccines to enhance the reactivity of the checkpoint blocking therapies, especially in the cohort of non-responding patients. The best clinical results for these therapies are observed in cancer types with high mutational load, such as melanoma [1, 2, 3], non-small-cell lung cancer [4], bladder cancer [5] and mismatch repair-deficient colorectal cancer [6••]. Recent studies highlighted the involvement of mutated neo-antigens as the main targets [7, 8, 9••, 10]. Clearly, subsets of patients bearing tumors with low mutational load may benefit as well, indicating that other immunogenic entities besides somatically mutated peptides are processed and presented to the immune cells, and these may be derived from shared cancer antigens, post-translationally modified peptides or translational products of new open reading frames that are unique to cancer.
In recent years, significant technological improvements in genomics and proteomics along with supportive bio-informatics and in silico prediction tools have facilitated major breakthroughs in the discovery of shared and neo-antigens that are presented as HLAp and are targets of anti-cancer T cell responses. Each having inherent advantages and disadvantages, both the in silico based prediction as well as the direct mass spectrometry (MS) based approaches led to successful identification of neo-antigens and provide a powerful toolbox for the development of cancer immunotherapy.
Section snippets
MS-based immunopeptidomics for discovery of tumor-associated antigens
The immunopeptidome is the collection of peptides presented by the HLA, also known as major histocompatibility complex (MHC). The human forms of the complex, HLA class I and HLA class II, are distinguished by the type of cells that express them, their intracellular processing and loading, and by the type of T cells that recognize them [11]. Currently, MS is the only unbiased methodology to comprehensively interrogate the repertoire of HLAp presented naturally in vivo [12•]. Of note, human
In silico approaches to identify tumor neo-antigens
Comprehensive lists of peptide sequences eluted from different HLA alleles together with affinity data from in vitro binding assays have been used to train algorithms that are now capable to predict potential HLAp that may be derived from any sequence of interest, including mutated genes [62, 63, 64, 65, 66, 67, 68]. For neo-antigens discovery, this in silico strategy uses as input mutation calls from massive parallel DNA and potentially also RNA sequencing data, comparing matched tumor and
MS approaches to identify tumor neo-antigens
Tumor associated antigens have been regularly discovered in the last two decades by MS; however it has been questioned whether MS would be sensitive enough to detect neo-antigens, especially as a discovery approach. Theoretically speaking, when the MS data is searched against customized personalized protein sequence databases with commonly used search engines like Mascot [85] or MaxQuant [86], the identification of novel private peptides should be computationally straightforward. Limitations
Integration of in silico and immunopeptidomics for high-throughput neo-antigens discovery
Both immunopeptidomics and in silico methodologies utilize similar mutation calling algorithms, but the thresholds applied to generate the customized references may be substantially different. In proteogenomics, permissive filters for calling mutations may be used to generate comprehensive references, and a strict false discovery rate threshold is then applied downstream at the level of MS-based detection of the peptides [88]. Due to limited sensitivity, false negatives are expected for
Conclusions
Not only has the direct biochemical approach revealed relevant targets for cancer immunotherapy, it has also contributed significantly to fundamental research related to antigen processing and presentation [92••]. Immunopeptidomics data in combination with cellular assays should be used to optimize predictors to score for immunogenicity. Predictors ought to credit peptide sequences that have been detected as true HLAp in immunopeptidomics studies, and take into account other characteristics of
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest
Acknowledgements
Supported by the Ludwig Institute for Cancer Research (LICR) (M.B.S and G.C.) and by the National Institute of Health Grant No. Ro1-CA156695, European Research Council Grant No. 1400206AdG-322875, the Leenaards Foundation and the Marcus Foundation (G.C.).
References (92)
- et al.
Integrated functional genomics approach for the design of patient-individual antitumor vaccines
Cancer Res
(2002) - et al.
The Tubingen approach: identification, selection, and validation of tumor-associated HLA peptides for cancer therapy
Cancer Immunol Immunother
(2004) - et al.
Expanding the detectable HLA peptide repertoire using electron-transfer/higher-energy collision dissociation (EThcD)
Proc Natl Acad Sci U S A
(2014) - et al.
Naturally occurring peptides associated with HLA-A2 in ovarian cancer cell lines identified by mass spectrometry are targets of HLA-A2-restricted cytotoxic T cells
Int Immunol
(2003) - et al.
Gene therapy with human and mouse T-cell receptors mediates cancer regression and targets normal tissues expressing cognate antigen
Blood
(2009) - et al.
Complementary IMAC enrichment methods for HLA-associated phosphopeptide identification by mass spectrometry
Nat Protoc
(2015) - et al.
MHC class I-associated phosphopeptides are the targets of memory-like immunity in leukemia
Sci Transl Med
(2013) - et al.
Extended O-GlcNAc on HLA class-I-bound peptides
J Am Chem Soc
(2015) - et al.
N-glycosylation enhances presentation of a MHC class I-restricted epitope from tyrosinase
J Immunol
(2009) - et al.
NetMHC-3.0: accurate web accessible predictions of human, mouse and monkey MHC class I affinities for peptides of length 8-11
Nucleic Acids Res
(2008)