Transcriptome profiles of carcinoma-in-situ and invasive non-small cell lung cancer as revealed by SAGE

Abstract

Background: Non-small cell lung cancer (NSCLC) presents as a progressive disease spanning precancerous, preinvasive, locally invasive, and metastatic lesions. Identification of biological pathways reflective of these progressive stages, and aberrantly expressed genes associated with these pathways, would conceivably enhance therapeutic approaches to this devastating disease.

Methodology/principal findings: Through the construction and analysis of SAGE libraries, we have determined transcriptome profiles for preinvasive carcinoma-in-situ (CIS) and invasive squamous cell carcinoma (SCC) of the lung, and compared these with expression profiles generated from both bronchial epithelium, and precancerous metaplastic and dysplastic lesions using Ingenuity Pathway Analysis. Expression of genes associated with epidermal development, and loss of expression of genes associated with mucociliary biology, are predominant features of CIS, largely shared with precancerous lesions. Additionally, expression of genes associated with xenobiotic metabolism/detoxification is a notable feature of CIS, and is largely maintained in invasive cancer. Genes related to tissue fibrosis and acute phase immune response are characteristic of the invasive SCC phenotype. Moreover, the data presented here suggests that tissue remodeling/fibrosis is initiated at the early stages of CIS. Additionally, this study indicates that alteration in copy-number status represents a plausible mechanism for differential gene expression in CIS and invasive SCC.

Conclusions/significance: This study is the first report of large-scale expression profiling of CIS of the lung. Unbiased expression profiling of these preinvasive and invasive lesions provides a platform for further investigations into the molecular genetic events relevant to early stages of squamous NSCLC development. Additionally, up-regulated genes detected at extreme differences between CIS and invasive cancer may have potential to serve as biomarkers for early detection.

Figure 1. Detection of carcinoma-in-situ bronchial lesions.

Bronchoscopy using A. white light for detection of CIS lesions (indicated by arrow), or B. LIFE (lung-imagine fluorescent endoscopy) for detection of CIS lesions (indicated by arrow). C. Histological section identifying a CIS lesion within the bronchial epithelium, typified by extensive squamous stratification.

Figure 2. Analysis of the top 300 most abundant tags from the BE, CIS, and invasive cancer datasets.

A. Cluster analysis of lung SAGE libraries. All SAGE libraries from this study, including five carcinoma in-situ libraries (CIS-1 through CIS-5), six invasive squamous cell carcinoma libraries (SCC-1 through SCC-6), one squamous metaplasia library (Met), and one squamous dysplasia library (Dys), as well as 14 bronchial epithelial libraries (BE-1 through BE-14), and two normal lung parenchyma SAGE libraries (LP-1, LP-2; accession GSE3708) generated in a previous study , , were analyzed by cluster analysis using an average-linkage algorithm. The top 300 most abundant tags were retained from each library, and analysis was based on 1128 unique tags in total. In the dendrogram, branch length represents distance. B–D. IPA functional analysis of the most abundant genes in the BE, CIS, and invasive cancer datasets. Tag-to-gene mappings for the top 300 most abundant tags from the BE, CIS, and SCC datasets, were used for IPA core analysis, consisting of 220, 231, and 233 IPA eligible mapped IDs, respectively. The three sets of data were displayed together using IPA core comparisons, and the five most significant functions within Physiological System Development and Function are shown for each of the three datasets. The data in B is sorted according to highest significance in BE, the data in C is sorted according to highest significance in CIS, and the data in D is sorted according to highest significance in invasive SCC. The orange line indicates the threshold limit of significance, preset at a p-value of 0.05. For a complete listing of the tags/mapped IDs used for this analysis see Table S1.

Figure 3. Venn diagrams of differentially expressed genes discussed in this manuscript.

Criteria for differential gene expression was defined as a minimal three-fold difference in normalized mean tag counts, and with a minimal mean tag abundance of 40 TPM in the over-expressing datasets. Up-arrows indicate up-regulated gene expression changes; down-arrows indicate down-regulated gene expression changes; numerical values refer to the number of differentially expressed tags. Areas of interception reflect gene expression changes in common between the two datasets. A. Expression changes in carcinoma-in-situ and precancerous lesions relative to BE. B. Expression changes in the cancer datasets relative to both bronchial epithelium and precancerous datasets. BE: bronchial epithelial; PC: precancer; CIS: carcinoma-in-situ; SCC: invasive squamous cell carcinoma.

Figure 4. IPA pathway graphical representation for the CIS_PC over BE dataset of up-regulated genes.

155 genes (IPA mapped IDs) are represented out of 190 SAGE tags up-regulated in both CIS and PC relative to BE. (See Table S2 for tag data.) Gene products are positioned according to subcellular localization. Only direct connections (i.e., direct physical contact between two molecules) among the individual gene products are shown for clarity of presentation; lines indicate protein-protein binding interactions, and arrows refer to “acts on” interactions such as proteolysis, expression, and protein-DNA/RNA interactions. Genes associated with epidermal development (see Table 3) are highlighted.

Figure 5. Genes up-regulated in the CIS and invasive SCC datasets relative to BE and PC.

A. Venn diagram of up-regulated SAGE tags and corresponding IPA mapped IDs for the CIS and SCC datasets. (See Table S5 and Table S6 for description of up-regulated tags in the CIS and SCC datasets, respectively.) B. IPA pathway graphical representation for the CIS over BE_PC dataset (80 unique IDs displayed in green; 58 shared IDs displayed in gray), and the SCC over BE_PC dataset (112 unique IDs displayed in red; 58 shared IDs displayed in gray). Gene products are positioned according to subcellular localization. Only direct connections (i.e., direct physical contact between two molecules) among the individual gene products are shown for clarity of presentation; lines indicate protein-protein binding interactions, and arrows refer to “acts on” interactions such as proteolysis, expression, and protein-DNA/RNA interactions. Eleven genes were detected at levels 20-fold or greater in the CIS over BE_PC dataset relative to the invasive cancer dataset (indicated by dark green), and 10 genes were detected at levels 20-fold or greater in the SCC over BE_PC dataset relative to the CIS dataset (indicated by dark red).

Figure 6. IPA canonical pathways analysis and toxicity lists analysis of the CIS over BE_PC and the SCC over BE_PC datasets.

For analysis of the CIS over BE_PC dataset, 109 IPA mapped IDs were eligible; for analysis of the SCC over BE_PC dataset, 153 IPA mapped IDs were eligible. The two sets of data were displayed together using IPA core comparisons, and the 10 most significant functions within Canonical Pathways and Toxicity Lists are shown above for each dataset. The data in A and C is sorted according to highest significance in CIS over BE_PC, and the data in B and D is sorted according to highest significance in SCC over BE_PC. The orange line indicates the threshold limit of significance, preset at a p-value of 0.05.

Figure 7. Correlation between up-regulated gene expression in CIS and SCC relative to BE and PC, with regions of frequent copy-number gain in CIS specimens.

Up-regulated genes (x-axis), plotted according to chromosomal location as indicated, were matched with segmental copy-number status (y-axis), defined by frequent copy-number gain (blue) and loss (red), from 20 independent CIS specimens. 224 genes were analyzed, and only those associated with regions gained at a minimal frequency of 0.2 are shown above. Knowledge of losses in addition to gains serves as a filter to identify those chromosomal regions that are preferentially gained rather than a reflection of general instability. See Table S12 for raw data pertaining to these analyses.

Figure 8. Correlation between down-regulated gene expression in CIS and SCC relative to BE and PC, with regions of frequent copy-number loss in CIS specimens.

Down-regulated genes (x-axis), plotted according to chromosomal location as indicated, were matched with segmental copy-number status (y-axis), defined by frequent copy-number loss (red) and gain (blue), from 20 independent CIS specimens. 81 genes were analyzed, and only those associated with regions lost at a minimal frequency of 0.2 are shown above. Knowledge of gains in addition to losses serves as a filter to identify those chromosomal regions that are preferentially lost rather than a reflection of general instability. See Table S13 for raw data pertaining to these analyses.