Integrated multi-omics landscape of non-small cell lung cancer with distant metastasis

Abstract

Background: Distant metastasis is one of the important factors affecting the prognosis of lung cancer patients. Extracellular vesicles (EVs) play an important role in the occurrence, development, and metastasis of cancer. However, it is currently unclear whether EVs in BALF are involved in distant tumor metastasis.

Methods: we collected bronchoalveolar lavage fluid (BALF) from patients with metastatic and non-metastatic non-small cell lung cancer (NSCLC) to isolate exosomes, which were then characterized by nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM), followed by comprehensive metabolomic and proteomic analysis to ultimately construct a distant metastasis prediction model for non-small cell lung cancer.

Results: Our research has found that the BALF of NSCLC patients is rich in EVs, which have typical morphology and size. There are significant differences in protein expression and metabolite types between patients with distant metastasis and those without distant metastasis. Sphingolipid metabolism pathways may be a key factor influencing distant metastasis in NSCLC. Subsequently, we constructed a predictive model for distant metastasis in NSCLC based on differentially expressed proteins identified by proteomics. This model has been proven to have high predictive value.

Conclusion: The multi-omic analysis generated in this study provided a global overview of the molecular changes, which may provide useful insight into the therapy and prognosis of NSCLC metastasis.

Keywords: exosome; lung cancer; metabolomic; metastasis; proteomics.

Figure 1

Overview of research design and characteristics of EVs. (A) The workflow of proteomics and omics experiments for BALF samples. (B) Typical TEM images of EVs in BALF (white arrows). Scale bar: 200 nm. (C) Using NAT to obtain typical images of the particle size distribution and concentration of EVs in BALF.

Figure 2

Proteomic changes and biological pathways of EVs in BALF of NSCLC patients. (A) Quantity of EVs proteins in BALF of each patient. (B) Differences in protein abundance between two groups of EVs. (C) PCA analysis of EVs proteins in distant metastasis group and non-distant metastasis group of lung cancer. (D) PLS-DA analysis of EVs proteins in group A and group B (E) Volcano plot shows the differential proteins between the group A and group B, each dot represents a protein, with red dots for proteins significantly upregulated in brain metastasis group and blue dots for proteins significantly downregulated in brain metastasis group. Fold change ≥ 2.0 and P < 0.05 is considered statistically significant. (F) KEGG pathway analysis of differential proteins.

Figure 3

Metabolomic changes of EVs in BALF of NSCLC patients. (A) PLS analysis of EVs metabolites in group A and group B (B) OPLS analysis of EVs metabolites in group A and group B (C) The heatmap showed the 22 metabolites with significant changes. (D) The lollipop chart shows the 10 down regulated proteins and 1 up regulated protein with the most significant changes in the distant metastasis group of lung cancer. (E) KEGG pathway analysis of differential metabolites.

Figure 4

Biomarker panel for early prediction of distant metastasis in NSCLC patients. (A) Constructing LASSO regression analysis model using differential proteins. (B) Prediction model of biomarkers for distant metastasis of lung cancer containing 15 proteins. (C–Q) Receiver operating characteristic curve (ROC) of AGAL, DCIL2, TPD54, CBL, PML, PCP, ASGL1, LEMD2, UBA5, PLCL2, CK052, H1.5, CEAM6, A1AG2 and FAK1. (R) The overall predictive performance of protein models. 95% confidence interval displayed on the graph.