The research team led by Nuno Barbosa Morais of iMM computationally analysed the expression of marker genes associated with a "fingerprint" of cancer cells in thousands of tumours revealing its therapeutic potential in the fight against cancer. The study published in this month's open access scientific journal PLoS Computational Biology * shows the types of tumours in which these genes are most active and identifies drugs with the potential to selectively eliminate cells under that label.

The centrosome is an organelle present in all animal cells, fundamental in several cellular processes, such as division, migration and communication between cells. It has been proposed for over a century that the abnormal increase in the number of these structures could induce cancer. Since then the increase in the number of centrosomes is seen as one of the "fingerprints" of cancer, and it gained scientists' attention. The technical difficulties in characterising this abnormality in patient samples has prevented its clinical potential from being explored on a larger scale. To avoid them, the team led by Nuno Barbosa Morais used the expression of genes causing this increase and analysed its incidence in thousands of tumours of different types of cancer and in normal tissue samples from the same patients. "The results revealed that this signature is present only in tumour samples, being more prevalent in aggressive forms of breast cancer", explains Nuno Barbosa Morais, adding "more importantly, the high number of these genes is associated with a lower survival rate in different types of cancer".



Using drug sensitivity data, researchers also identified selective compounds for cells with this abnormality that could be targeted specifically against cancer cells, not affecting patients' healthy cells.
"In addition, the samples we analysed are characterised by their DNA sequence and the expression of thousands of genes, so the integration of this data allowed us to understand better the causes and molecular consequences of this increase of centrosomes in our cells", explains Bernardo de Almeida, the first author of this study.

"The next steps are to translate the expression data of the genes that cause this "fingerprint" into clinical decision support information. We also intend to validate the effectiveness of the drugs identified by our computational approach as having greater therapeutic potential. These studies will naturally involve collaborations with colleagues specialised in clinical oncology and pharmacology," says Nuno Barbosa Morais, laboratory lead and research supervisor.

This study involved collaboration with Joana Paredes' team at i3S - Institute of Research and Innovation of Health in Porto and Mónica Bettencourt Dias from the Gulbenkian Institute of Science. The work was funded by the European Molecular Biology Organisation (EMBO) and the Foundation for Science and Technology (FCT).



*Bernardo P. de Almeida, Andre F. Vieira, Joana Paredes, Mónica Bettencourt-Dias, Nuno L. Barbosa-Morais (2019) Pan-cancer association of a centrosome amplification gene expression signature with genomic alterations and clinical outcome. PLoS Computational Biology. https://doi.org/10.1371/journal.pcbi.1006832

size="30"

Inês Domingues

Director of iMM Communications