In-Depth Characterization of Mass Spectrometry-Based Proteomic Profiles Revealed Novel Signature Proteins Associated with Liver Metastatic Colorectal Cancers

Abstract

Liver metastasis is the most common form of metastatic colorectal cancers during the course of the disease. The global change in protein abundance in liver metastatic colorectal cancers and its role in metastasis establishment have not been comprehensively analyzed. In the present study, fresh-frozen tissue samples including normal colon/localized/liver metastatic CRCs from each recruited patient were analyzed by quantitative proteomics using a multiplexed TMT labeling strategy. Around 5000 protein groups were quantified from all samples. The proteomic profile of localized/metastatic CRCs varied greatly from that of normal colon tissues; differential proteins were mainly from extracellular regions and participate in immune activities, which is crucial for the chronic inflammation signaling pathways in the tumor microenvironment. Further statistical analysis revealed 47 proteins exhibiting statistical significance between localized and metastatic CRCs, of which FILI1P1 and PLG were identified for the first time in proteomic data, which were highly associated with liver metastasis in CRCs.

Figure 1

Workflow and multiplexed labeling strategy of this study. (a) Proteomic workflow of this study. (b) Labeling strategy applied in this study: peptides were labeled with TMT 10plex reagents. All tissue samples were split into three labeling batches (group 1/2/3). First, the pooled reference sample was generated by pooling aliquots of each individual sample from all patients, which was further assigned in the TMT labeling experiment as the reference channel. The 9 serial samples (T for tumor, M for metastasis, and N for normal colon) and the reference sample were together included in one TMT labeling experiment group, and there are 3 groups in total.

Figure 2

Overview of patient proteomic profiles, including normal colon (N), localized (T), and metastatic (M) CRC tissue types, as labeled at the end of each sample name. (a) Venn diagram of protein identifications from three different tissue types. (b) Principal component analysis (PCA) on the identified proteins from all samples revealed significant difference between normal colon (N) and localized (T)/metastatic (M) CRC. (c) Hierarchical cluster analysis on differentially expressed proteins among all samples. (d) Proteins with top significance (fold change > 2) were prioritized and annotated. Most of these proteins were from extracellular regions and participate in immune activities. (e) Correlation analysis of all samples. Results showed that proteomic profiles within normal tissue (N) samples were highly correlated. However, very poor correlation was found among metastatic samples which indicated the high heterogeneity of metastatic CRC.

Figure 3

Differentially expressed proteins between different sample groups. For normal colon (N) and localized CRCs (T), volcano plot (a) revealed 66 significant proteins; these proteins were further clustered and presented in heat map (b), with 38 upregulated and 28 downregulated proteins in T (principal component analysis). (c) Using identified significant proteins showed very good separating power on PC1 dimension to distinguish normal (N) and cancer (T) tissues. Similar analysis was performed between normal colon (N) and metastatic CRC (M), with 120 significant proteins prioritized from volcano plot (d), in which 46 proteins were upregulated and 74 proteins were downregulated in M (e); very good separation between N and M on PC2 dimension was achieved using 120 significant proteins.

Figure 4

Proteins to differentiate localized (T) and metastatic (M) CRCs. Statistical analysis revealed 47 differentially expressed proteins. These proteins were presented using clustering analysis (a) and principal component analysis (b). The expression of selected significant protein candidates among different groups, e.g., normal, tumor (T), and metastasis (M), was provided in c and d, in which the y-axis showed the relative intensities of the proteins quantified in each tissue group.