SAPrIm, a semi-automated protocol for mid-throughput immunopeptidomics

Abstract

Human leukocyte antigen (HLA) molecules play a crucial role in directing adaptive immune responses based on the nature of their peptide ligands, collectively coined the immunopeptidome. As such, the study of HLA molecules has been of major interest in the development of cancer immunotherapies such as vaccines and T-cell therapies. Hence, a comprehensive understanding and profiling of the immunopeptidome is required to foster the growth of these personalised solutions. We herein describe SAPrIm, an Immunopeptidomics tool for the Mid-Throughput era. This is a semi-automated workflow involving the KingFisher platform to isolate immunopeptidomes using anti-HLA antibodies coupled to a hyper-porous magnetic protein A microbead, a variable window data independent acquisition (DIA) method and the ability to run up to 12 samples in parallel. Using this workflow, we were able to concordantly identify and quantify ~400 - 13000 unique peptides from 5e5 - 5e7 cells, respectively. Overall, we propose that the application of this workflow will be crucial for the future of immunopeptidome profiling, especially for mid-size cohorts and comparative immunopeptidomics studies.

Keywords: DIA; HLA-bound peptides; cancer antigen; human leukocyte antigen; immunopeptidomics.

Figure 1

HLA class I and II antigen-presentation pathways. The HLA class I pathway (above) is responsible for degrading endogenous antigens into peptides via the multi-catalytic proteasome complex. Peptides are then transported into the endoplasmic reticulum (ER) by the transporter associated with antigen processing (TAP). Here peptides bind based on their relative affinity to the allotypes present. HLA-I peptide complex is then transported through the Golgi apparatus to the cell surface where it is scrutinised by CD8+ T cells. The HLA class II pathway (below) involves the degradation of exogenous antigens in the endosome compartment. In the ER immature HLA-II proteins are stabilised by the invariant chain and released into the HLA class II compartment. Here, the enzyme HLA-DM removes the class II invariant chain peptide (CLIP) from the binding pocket allowing for antigenic peptide binding. Mature HLA-II molecules bound to their peptide cargo are then transported to the cell surface for CD4+ T cell recognition.

Figure 2

Diagram showing reagents plating format for the KingFisher. Volumes as well as the KingFisher method timing is specified in each row.

Figure 3

Diagram showing reagents plating format for the C18 stage tip clean up in a Lobind 96 well plate. This step is manual and not performed by KingFisher Duo. Conditioning buffer: 50%ACN/0.1% TFA. Wash buffer: 0.1% TFA. HLA-I Elution Buffer: 28% ACN/ 0.1% TFA. HLA-II Elution Buffer: 32% ACN/ 0.1% TFA.

Figure 4

The average number of peptides per cell count pellet. The number of peptides identified for each sample were as follows: 1e5 (black), 5e5 (purple), 5e6 (blue), 1e7 (green), and 5e7 (orange). Please note that the experiment for the 1e5 samples was performed separately to evaluate the limit of detection. Statistical significance with a p-value < 0.0001, as determined by an one-way ANOVAa t-test, is denoted by ****.

Figure 5

Length distribution of peptides bound to HLA class I. The x-axis shows the length of the peptides for each condition and the y-axis the percentage frequency for each length.

Figure 6

Percentage of predicted binders versus non-binders. Bar graph showing the percentage of binders for each condition following binding predictions on NetMHCpan4.1 (16).

Figure 7

The overlap between peptides identified across different cell counts. The overlap of unique peptides across different cell counts (5e5 in purple, 5e6 in blue, 1e7 in green, and 5e7 in orange) is represented in a Venn diagram.

Figure 8

Violin plot depicting the peptide intensity between commonly identified HLA peptides. Violin plot showing the trend in intensities for the 521 overlapping peptides across four conditions. The mean log2 intensities for each condition are indicated by the values. Statistical significance with a p-value < 0.0001, as determined by an one-way ANOVA, is denoted by ****.