Small noncoding RNAs can be used to predict if a person has breast cancer, concluded researchers who contributed to The Cancer Genome Atlas (TCGA) project. The results indicate that differences in the levels of specific types of noncoding RNAs can be used to distinguish between cancerous and noncancerous tissues. These RNAs can also be used to classify cancer patients into subgroups of persons that have different survival outcomes.
Small noncoding RNAs are RNA molecules that do not give rise to proteins but which may have other important functions in the cell. Colead author Steven Jones, PhD, of Simon Fraser University and the University of British Columbia in Vancouver, explained, “For many years, small noncoding RNAs near transcriptional start sites have been regarded as ‘transcriptional noise’ due to their apparent chaotic distribution and an inability to correlate these molecules with known functions or disease.”
Jones added that TCGA generated small RNA sequence information, and that the research team used a computational approach for analysis to find clinically useful information. Their experiments, which were published in EMBO Reports (2014; doi:10.1002/embr.201337950), indicated that genome-wide changes in the expression levels of small noncoding RNAs in the first exons of protein-coding genes are associated with breast cancer.
Many different small noncoding RNAs were able to be distinguished between healthy persons and those with breast cancer. These RNA molecules were mapped to specific locations on the DNA sequence. Then, the researchers looked for correlations between the noncoding RNAs that were strongly expressed and the disease status of the patients from whom the tissue samples were isolated.
The scientists were able to distinguish between the many different small noncoding RNAs that are found near the transcriptional start sites of genes in healthy persons and patients with breast cancer (in this case, breast invasive carcinoma). The researchers then tested if the expression of the small RNAs in the identified genomic locations could predict the presence of disease in another group of tissue samples from patients known to have breast cancer. The test efficiently predicted the correct disease status for the samples in the new study group.
“The potential to predict cancer status is restricted to only a subset of the many small noncoding RNAs found near transcription start sites of the genes. What’s more, these RNA locations are highly enriched with CpG islands,” said first author Athanasios Zovoilis, MD, PhD, of Harvard University in Cambridge, Massachusetts. CpG islands are genomic regions that contain a high frequency of cytosine and guanine. The presence of these RNAs in these islands may implicate their involvement with DNA methylation processes and the onset of disease but additional experiments are needed to explore and prove this link.
The generation of data by TCGA, which now provides access to large amounts of sequencing information for diseased and normal tissues, made the work possible. TCGA is now one of the largest resources for small noncoding RNAs in existence.