Colorectal cancer : genome, transcriptone and proteome dynamics

Abstract

Colorectal carcinomas are the third most common malignancies in the Western World. Despite screening programs about 70% of these tumors are detected at advanced stages. Late diagnosis, however, results in a significant reduction of average survival times. Thus, alternative tools or biomarkers for early detection, prognosis and therapeutic intervention are needed.

Patients with ulcerative colitis (UC) have an increased risk to develop an ulcerative colitisassociated colorectal carcinoma (UCC). To evaluate independent cellular features with possible predictive value, a total of 683 UC biopsies from two patient groups were analyzed by image cytometry and immunohistochemistry: group A consisted of 8 patients with UCC, group B of 16 ulcerative colitis patients without any malignancy. Whereas no difference in inflammatory activity and dysplasia between patient groups could be observed, genomic aneuploidy, laminin-5 immunopositivity and increased Cyclin A expression could be confirmed as significant predictors for malignant transformation.

We then analyzed 23 UC associated colorectal carcinomas by image cytometry, comparative genomic hybridization (CGH), and immunohistochemistry. We found that DNA aneuploidy in UCC correlates with specific chromosomal imbalances and increased cyclin A expression levels. Distinct amplifications that were not known to be associated with colorectal cancer will likely reveal novel genes involved in colorectal tumorigenesis. However, the overall pattern of specific chromosomal aberrations in UCC is similar to that seen in sporadic colorectal carcinomas.

To further elucidate the importance of these highly conserved chromosomal aberrations, we modeled specific chromosomal ancuploidies in cancer cells and dissected the immediate consequences of genomic imbalances on the transcriptome: we generated artificial trisomies in a diploid colorectal cancer cell line and in normal human breast epithelial cells using microcell-mediated chromosome transfer. Gene expression levels were analyzed using cDNA arrays. Regardless of chromosome or cell type, chromosomal trisomies resulted in a significant increase in the average transcriptional activity of the trisomic chromosome and affected the expression of numerous genes on other chromosomes as well, revealing a more complex global transcriptional dysregulation.

Analogous results could be obtained in primary colorectal tumors: we found that specific and recurrent chromosomal aneuploidies exert strong and direct influence on gene expression levels of resident genes on the affected chromosomes by analyzing tissue samples from 36 patients with sporadic colorectal carcinoma. In addition, increasing genomic instability, aneuploidy and a recurrent pattern of chromosomal aberrations are accompanied by distinct gene- and protein expression patterns that correlate with subsequent stages of colorectal cancer progression.

Our analysis identified 58 genes differentially expressed between normal mucosa and adenoma, 116 genes between adenoma and carcinoma, and 158 genes between primary carcinoma and liver metastasis. Proteomics analysis revealed 42 differentially expressed proteins that were related to similar canonical pathways as the differentially expressed genes. The identified proteins underwent extensive posttranslational modifications, thus multiplying the transcriptional dysregulation. A malignancy related protein profile could also be identified in serum samples: comprehensive serum protein profiling in 147 patients with colorectal malignancy and healthy individuals using SELDI-TOF mass spectrometry revealed a set of 13 differentially expressed features. Using 16 different classifiers allowed class prediction with 96.7% sensitivity and 100% specificity in a blinded validation set, including patient sera with early stage disease. Large scale prospective multicenter studies are now warranted to establish the clinical value of SELDI-TOF based protein profiling for colorectal cancer screening.