Glycoproteomics Analysis of Triple Wild-Type Lung Adenocarcinoma Tissue Samples

Abstract

Lung cancer has both high incidence and mortality, making it the leading cause of cancer-related mortality worldwide. It is a highly heterogeneous disease, with several histological subtypes and genetic alterations that influence prognosis and available treatment options. Here, we focus on the triple wild-type (TWT) subtype of lung adenocarcinoma (LUAD) that lacks the three most common actionable genetic alterations, subsequently making targeted therapies inaccessible. In this study, our aim was the mass spectrometry-based proteomic and N-glycoproteomic characterization of tumor and adjacent normal lung tissue regions from individuals (n = 12) with TWT LUAD. We found several proteins previously identified as potential prognostic or diagnostic biomarkers in LUAD and described dysregulated biological processes, giving an overview of the general differences between healthy and tumor tissue. Also, we highlight specific signatures detected using N-glycoproteomics and discuss their potential and importance based on data from databases and literature. To the best of our knowledge, this is the first N-glycoproteomics-focused study on TWT LUAD, and it could provide a valuable resource for further studies into this less well characterized subtype of lung cancer. For instance, we report altered N-glycosylation for several glycoproteins implicated in LUAD and other cancers that could have functional importance connected to the disease.

Keywords: Cancer research; FFPE tissue; Lung adenocarcinoma; Mass spectrometry; N-Glycoproteomics.

Figure 1

Analysis workflow.

Figure 2

(A) Principal component analysis of tumor adjacent normal (red) and tumor (blue) samples. (B) Sample-wise correlation heatmap of tumor adjacent normal (red) and tumor (blue) samples.

Figure 3

Volcano plot of protein Fold-Change values (x-axis, log 2 transformed) vs adjusted p-values (y-axis). Proteins overexpressed in tumor regions compared to adjacent normal are red; those underexpressed in tumor regions compared to adjacent normal are blue. Proteins with not statistically significant changes (below the 0.05 FDR threshold) are gray.

Figure 4

(A) Treeplot of the top 20 enriched GOBP terms based on adjusted p-values. Clustering is based on the overlap between gene sets, nodes are colored based on Normalized Enrichment Scores (NESs), and the size of the nodes represents the number of genes in the gene set. The names of the individual clusters are based on the gene sets within. (B) Enrichment map of the top 75 enriched GOBP terms based on adjusted p-values. Edges represent the overlap between data sets, nodes are colored based on NESs, and the size of the nodes represents the number of genes in the gene set.

Figure 5

Overall N-glycosylation metrics for the tumor and tumor adjacent tissue. Metrics were weighted with N-glycopeptide expression and averaged across all of the N-glycopeptides. (A) Overall sialylation and an example N-glycopeptide with two antennae, one with sialic acid and one without—a sialylation of 0.5. (B) Overall galactosylation and an example N-glycopeptide with two antennae, both containing galactose units—a galactosylation of 1.0. (C) Overall fucosylation and an example N-glycan with core fucosylation—a fucosylation of 1.0.

Figure 6

Changes in N-glycopeptide abundance between tumor adjacent normal and tumor samples (red and blue bars, respectively), and N-glycoprotein fold-change values (black diamonds). The secondary y-axis (from 0 to 12) shows protein fold-changes, and the dotted line at 1 shows the threshold for underexpression (fold-change under 1) and overexpression (fold-change over 1) in tumor regions. The nomenclature used for glycopeptides includes the UniProt short name for the protein, the location of the N-glycosylated asparagine in the sequence, and the attached glycan. The glycan name includes the number of fucose (F), hexose (H), N-acetylhexosamine (N), and sialic acid (S) residues.