Email updates

Keep up to date with the latest news and content from BMC Cancer and BioMed Central.

Open Access Research article

Signs of positive selection of somatic mutations in human cancers detected by EST sequence analysis

Vladimir N Babenko1, Malay K Basu1, Fyodor A Kondrashov2, Igor B Rogozin1 and Eugene V Koonin1*

Author Affiliations

1 National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda MD, USA

2 Section of Ecology, Behavior and Evolution, University of California San Diego, La Jolla, CA, USA

For all author emails, please log on.

BMC Cancer 2006, 6:36  doi:10.1186/1471-2407-6-36

Published: 9 February 2006

Abstract

Background

Carcinogenesis typically involves multiple somatic mutations in caretaker (DNA repair) and gatekeeper (tumor suppressors and oncogenes) genes. Analysis of mutation spectra of the tumor suppressor that is most commonly mutated in human cancers, p53, unexpectedly suggested that somatic evolution of the p53 gene during tumorigenesis is dominated by positive selection for gain of function. This conclusion is supported by accumulating experimental evidence of evolution of new functions of p53 in tumors. These findings prompted a genome-wide analysis of possible positive selection during tumor evolution.

Methods

A comprehensive analysis of probable somatic mutations in the sequences of Expressed Sequence Tags (ESTs) from malignant tumors and normal tissues was performed in order to access the prevalence of positive selection in cancer evolution. For each EST, the numbers of synonymous and non-synonymous substitutions were calculated. In order to identify genes with a signature of positive selection in cancers, these numbers were compared to: i) expected numbers and ii) the numbers for the respective genes in the ESTs from normal tissues.

Results

We identified 112 genes with a signature of positive selection in cancers, i.e., a significantly elevated ratio of non-synonymous to synonymous substitutions, in tumors as compared to 37 such genes in an approximately equal-sized EST collection from normal tissues. A substantial fraction of the tumor-specific positive-selection candidates have experimentally demonstrated or strongly predicted links to cancer.

Conclusion

The results of EST analysis should be interpreted with extreme caution given the noise introduced by sequencing errors and undetected polymorphisms. Furthermore, an inherent limitation of EST analysis is that multiple mutations amenable to statistical analysis can be detected only in relatively highly expressed genes. Nevertheless, the present results suggest that positive selection might affect a substantial number of genes during tumorigenic somatic evolution.