Proteomic analysis enables distinction of early- versus advanced-stage lung adenocarcinomas

Abstract

Background: A gel-free proteomic approach was utilized to perform in-depth tissue protein profiling of lung adenocarcinoma (ADC) and normal lung tissues from early and advanced stages of the disease. The long-term goal of this study is to generate a large-scale, label-free proteomics dataset from histologically well-classified lung ADC that can be used to increase further our understanding of disease progression and aid in identifying novel biomarkers.

Methods and results: Cases of early-stage (I-II) and advanced-stage (III-IV) lung ADCs were selected and paired with normal lung tissues from 22 patients. The histologically and clinically stratified human primary lung ADCs were analyzed by liquid chromatography-tandem mass spectrometry. From the analysis of ADC and normal specimens, 4863 protein groups were identified. To examine the protein expression profile of ADC, a peak area-based quantitation method was used. In early- and advanced-stage ADC, 365 and 366 proteins were differentially expressed, respectively, between normal and tumor tissues (adjusted P-value < .01, fold change ≥ 4). A total of 155 proteins were dysregulated between early- and advanced-stage ADCs and 18 were suggested as early-specific stage ADC. In silico functional analysis of the upregulated proteins in both tumor groups revealed that most of the enriched pathways are involved in mRNA metabolism. Furthermore, the most overrepresented pathways in the proteins that were unique to ADC are related to mRNA metabolic processes.

Conclusions: Further analysis of these data may provide an insight into the molecular pathways involved in disease etiology and may lead to the identification of biomarker candidates and potential targets for therapy. Our study provides potential diagnostic biomarkers for lung ADC and novel stage-specific drug targets for rational intervention.

Keywords: clinical proteomics; lung adenocarcinomas; mass spectrometry; proteomics.

FIGURE 1

A, Representative H&E staining of tumor and normal adjacent tissue samples. It illustrates the representative tumor (right) and normal adjacent (left) tissue sections from early‐stage ADC (A) and (C), and from advanced‐stage ADC (B) and (D). B, Principal component analysis (PCA) of protein expression patterns of the 44 samples. The greatest variance (PC1_24.1%) is given by the differences between normal and tumor tissue and the second (PC2_ 9.5%) by differences between stages. C, Hierarchical clustering separated normal, early, and advanced ADC proteomes. Heat map and hierarchical clustering based on 1579 differentially expressed proteins (DEPs) from the ANOVA test. Heat map colors are based on the z‐scored (log₂) intensity values. Blue and red correspond to decreased and increased expression levels, respectively

FIGURE 2

A, Venn diagrams showing the distribution of the identified protein groups across the sample classifications. T refers to tumor samples. N refers to adjacent normal tissue sample. B, Gene ontology (GO) annotation of identified proteins by cellular component

FIGURE 3

Differentially expressed proteins in early‐stage ADC. A, Volcano plot of log₂‐fold change (FC) versus P‐value t‐test. Points alter in size according to the magnitude of the fold change. B, Heat map generated by hierarchical clustering of the most altered proteins. The heat map shows a clear separation of adjacent normal and tumor tissues (–4 ≥ log₂ FC ≥ 4, adj. P < .01). Blue and red correspond to decreased and increased expression levels, respectively

FIGURE 4

Comparison between normal and tumor tissue in advanced‐stage ADC. A, Volcano plot of log₂‐fold change (FC) versus P‐value t‐test. Points alter in size according to the magnitude of the fold change. B, Heat map generated by hierarchical clustering of the most altered proteins. The heat map shows a clear separation of adjacent normal and tumor tissues (–4 ≥ log₂ FC ≥ 4, adj. P < .01). Blue and red correspond to decreased and increased expression levels, respectively

FIGURE 5

Comparison of early‐ and advanced‐stages in tumor tissue. A, Volcano plot of log₂‐fold change (FC) versus P‐value t‐test. Points alter in size according to the magnitude of the fold change. B, Heat map generated by hierarchical clustering of the most altered proteins. The heat map shows a clear separation of adjacent normal and tumor tissues (–4 ≥ log₂ FC ≥ 4, adj. P < .01). Blue and red correspond to decreased and increased expression levels, respectively

FIGURE 6

Protein‐protein interaction and correlation for the 18 early‐stage ADC‐specific proteins. A, Protein network from STRING. Shown in red are the significant interactions (FDR < .05 with the highest confidence threshold). B, Spearman correlation. The color scale represents the strength of the correlation (r) (white to red, positive correlation; white to blue, negative correlation). *P < .05; **P < .01