Transcriptomic and proteomic characterization of cell and protein biomarkers of checkpoint inhibitor-induced liver injury

Abstract

Immune checkpoint inhibitors (ICI) targeting CTLA-4 and PD-1 have shown remarkable antitumor efficacy, but can also cause immune-related adverse events, including checkpoint inhibitor-induced liver injury (ChILI). This multi-omic study aimed to investigate changes in blood samples from treated cancer patients who developed ChILI. PBMCs were sequenced for by transcriptomic and T cell receptor repertoire (bulk and single-cell immune profiling), and extracellular vesicle (EV) enrichment from plasma was analyzed by mass spectroscopy proteomics. Data were analyzed by comparing the ChILI patient group to the control group who did not develop ChILI and by comparing the onset of ChILI to pre-ICI treatment baseline. We identified significant changes in T cell clonality, gene expression, and proteins in peripheral blood mononuclear cells (PBMCs) and plasma in response to liver injury. Onset of ChILI was accompanied by an increase in T cell clonality. Pathway analysis highlighted the involvement of innate and cellular immune responses, mitosis, pyroptosis, and oxidative stress. Single-cell RNA sequencing revealed that these changes were primarily found in select T cell subtypes (including CD8 + effector memory cells), while CD16 + monocytes exhibited enrichment in metabolic pathways. Proteomic analysis of plasma extracellular vesicles showed enrichment in liver-associated proteins among differentially expressed proteins. Interestingly, an increase in PBMC PD-L1 gene expression and plasma PD-L1 protein was also found to be associated with ChILI onset. These findings provide valuable insights into the immune and molecular mechanisms underlying ChILI as well as potential biomarkers of ChILI.Trial registration number NCT04476563.

Keywords: Biomarker; ChILI; Checkpoint inhibitor; Liver injury; PD-L1.

Fig. 1

Mixed prospective study experimental design. The various study groups are delineated by BC (before checkpoint inhibitor), AC (after checkpoint inhibitor control), TOL (time of liver injury), and FU (follow-up) timepoints 1 and 2. Under each group designation, the number of PBMC samples assessed for RNAseq is shown, followed by the number of plasma samples assessed by proteomics in parentheses. The numbers along the arrows indicate the number of patients in control and ChILI groups for whom BC samples were available. Additional ChILI subjects lacking BC samples were added to increase the power of the AC vs TOL comparison

Fig. 2

T cell receptor clonality. Gini index was calculated for each sample. The Gini indices for TCRA (A) and TCRB (B) sequences are shown over time for subjects that developed ChILI. *p < 0.05, ****p < 0.0001

Fig. 3

Heatmap showing the top 50 differentially expressed genes among samples from TOL, AC and BC. Each column represents a single sample. Red represents up-regulation, blue represents down-regulation. In the top dendrogram, light blue represents TOL samples, dark blue represents AC samples, and green represents BC samples

Fig. 4

Single-cell RNAseq. Number of DEGs in each of the cell types in PBMC comparing post-treatment to pre-treatment for those subjects that developed ChILI (A) and those subjects that did not develop ChILI (B)

Fig. 5

Results of proteomic analysis of EV-enriched plasma. A Venn diagram depicting the overlap between total proteins detected, quantified, and those annotated as enriched in liver. Volcano plots highlighting the enrichment of liver-associated DEPs for B TOL vs BC and C TOL vs AC. Red circles denote increased expression, blue circles denote decreased expression of liver-associated proteins