Enrichment and identification of phosphorylated and glycosylated MHC-I peptides from clinical tumor samples and murine models

Despite extensive research, cancer is the second cause of death globally and in the USA. In the past decade, the immune checkpoint blockade (ICB) therapies with antibodies that target T cell checkpoint protein have revolutionized cancer treatment. However, the ICB is not selective and tumor responses vary, suggesting the importance of identifying cancer neoantigens to develop targeted immunotherapies. Previous studies in the Hunt lab revealed that a deregulated cell signaling in cancer results in the presentation of class I Human leukocyte antigen (HLA-I) peptides with phosphorylation or O-linked β-N-acetylglucosamine (O-GlcNAc) modifications, which are potential targets for the T cell recognition.
We have investigated the presentation of HLA-I phosphopeptides from the clinical head and neck cancer (HNCs) samples and intrahepatic cholangiocarcinoma (iCCA) tumors. The investigation on the presentation of HLA-I phosphopeptides on HCC is also continued by the comparative analysis of tumor tissue and adjacent cirrhotic tissues. The overlapping of the identified HLA-I phosphopeptides among various cancer types and cancer patients shows their immunotherapeutic merits.
We have also investigated the presentation of MHC-I phosphopeptides in murine melanoma, lung, colon, and liver carcinoma cell lines. Results show that cancer in the murine model has a similar characteristic of shared MHC-I phosphopeptides among different cancer types like humans.
Mass spectrometry (MS) has been proved to be a powerful technique in the identification of HLA-I peptides. Different fragmentation techniques were used in this dissertation to analyze phosphorylated or glycosylated HLA-I peptides. Complementary data obtained from the collisional activated dissociation (CAD) and electron transfer dissociation (ETD) was used for the manual sequencing and phosphosite determination for HLA-I phosphopeptides. For glycosylated HLA-I peptides, high energy collisional activated dissociation (HCD), electron transfer high collision dissociation (EThcD), and ETD fragmentation methods were used for the identification of glycan type besides the peptide sequencing.
The identification of phosphorylated and glycosylated HLA-I peptides in the immunopeptidome of typical sub-gram clinical tumor samples requires their enrichment before the MS analysis. Current technology demands 3 to 5 E8 cell equivalents to identify sub-femtomole HLA-I phosphopeptide and glycopeptides. This requirement limits the number and type of clinical tumor samples analyzed for the post-translationally modified (PTM) HLA-I peptides.
Here, the sample size requirement restriction was addressed in two separate studies. First, we investigated the application of a murine xenograft model versus prolonged in vitro cell line culturing for the HLA-I phosphopeptide analysis. Results showed that human cell lines grown as xenografts in immunodeficient mice also presented equivalent HLA-I peptide and phosphopeptides repertoires to those grown in culture. Therefore, they can provide an alternate means to generate sufficient cell quantities for HLA-I phosphopeptide analysis from tumor biopsy samples in the future. Second, the sequential immobilized boronic acid enrichment (IBAE) in anhydrous condition was developed for the enrichment of glycosylated HLA-I peptides from the flowthrough of immobilized metal affinity chromatography (IMAC), which was implemented for phosphopeptide enrichment. We reported novel glycan modifications on HLA-I peptide samples isolated from clinical hepatocellular carcinoma (HCC) tumors at attomole levels, illustrating the introduced IBAE method's sensitivity.