EMT is the dominant program in human colon cancer

Area of Science:

• Oncology
• Molecular Biology
• Genomics

Background:

• Colon cancer prognosis is traditionally determined post-surgery using clinicopathologic features.
• Existing methods offer limited predictive power before invasive procedures.

Purpose of the Study:

• To identify intrinsic molecular subtypes of colon cancer using gene expression data.
• To determine if molecular subtypes can predict disease progression and recurrence.

Main Methods:

• Unsupervised analysis of microarray data from 326 colon cancer samples.
• Identification of the first principal component (PC1) representing the most variable gene expression patterns.
• Correlation analysis between PC1 and the epithelial-mesenchymal transition (EMT) signature.
• MicroRNA screening to identify regulators of PC1.

Main Results:

• PC1 identified two intrinsic molecular subtypes of colon cancer.
• PC1 strongly correlated with the EMT signature (Pearson R = 0.92, P < 10(-135)).
• MicroRNA-200 (MiR200) was identified as the most anti-correlated microRNA with PC1, known to regulate EMT.

Conclusions:

• Colon cancer biology can be clarified through transcriptome analysis, revealing intrinsic molecular subtypes.
• The dominant gene expression pattern (PC1) is linked to EMT, providing a molecular basis for classification.
• This approach offers a new perspective on the heterogeneity of colon cancer.