Identification of novel therapeutic targets by genome analysis in breast cancer

The survival of breast cancer patients has increased over the years due to earlier detection and more personalized treatments. Despite these advancements in cancer care, still many patients do not benefit from the given treatment. The mechanisms behind lack of treatment response or resistance to treatment are not yet fully understood. Albeit novel drugs are designed, the conversion from efficient drugs in functional systems to successful use in clinic is poor. This has increased the demand for more in vivo-like model systems for cancer studies. The aims of the project were to propose a more in vivo-like model system for use in cancer research, to identify gene-drug interactions as possible predictors for drug response in HER2+ breast cancers and to elucidate the effects of miRNAs on mRNA-protein translation in breast cancer. Both in vitro studies of established breast cancer cell lines, in vivo studies of xenografts as well as genomic and transcriptomic analyses of biopsies from breast cancer patients were performed. Jernström and colleagues reported growth inhibition in 2D and 3D culture models to be dependent on different compounds. In addition, comparisons of gene expression analyses revealed that two of the studied 3D models were similar to xenografts, suggesting these models to be more in vivo-like. Drug screening of 13 HER2+ cell lines showed that drug responses depend on mutation status of PIK3CA and PTEN, two known factors for trastuzumab resistance. From elastic net analysis the MAPK8 gene was suggested as a predictor for response to the Akt1/2 kinase inhibitor in our library, to which two trastuzumab resistant cell lines responded to. This indicates that HER2 protein level is not enough when deciding treatment, but other factors should be taken into account as well. In the third study, patient material from a breast cancer cohort was used for integration of mRNA, miRNA and protein expression data to study interactions on the translational level. This lead to an interactome map describing the effects of miRNAs on mRNA-protein translation. Not only did miRNAs lead to deprivation of protein translation, but it also indirectly functioned as an enhancer. In addition, many proteins seem to be affected by several miRNAs in both positive and negative direction. This study contributes with knowledge to the still not fully understood mechanisms of miRNAs. The thesis was delivered in May 2014, and defended the 11th of December 2014 at Medical Faculty at University of Oslo. Publications: Vesa Hongisto*, Sandra Jernström*, Vidal Fey, John Patrick Mpindi, Kristine Kleivi Sahlberg, Olli Kallioniemi, Merja Perälä; High-throughput 3D screening reveals differences in drug sensitivities between culture models of JIMT1 breast cancer cells; PLoS One, 23rd October 2013, vol.8, no.10 Sandra Nyberg, Vesa Hongisto, Suvi-Katri Leivonen, Dagim Shiferaw Tadele, Henrik Edgren, Olli Kallioniemi, Merja Perälä, Gunhild Mari Mælandsmo, Anne-Lise Børresen-Dale, Kristine Kleivi Sahlberg; Identifying alternative compounds for HER2 positive breast cancer cells by drug screening; Submitted. Integrated analysis reveals microRNA networks coordinately expressed with key proteins in breast cancer. Miriam Ragle Aure†, Sandra Jernström†, Marit Krohn, Hans Kristian Moen Vollan, Eldri U Due, Einar Rødland, OSBREAC, Rolf Kåresen, Prahlad Ram, Yiling Lu, Gordon B Mills, Kristine Kleivi Sahlberg, Anne-Lise Børresen-Dale, Ole Christian Lingjærde, Vessela N Kristensen. In press, Genome Medicine 2014. † Shared first authors. This study is of importance for future breast cancer patients. Many HER2+ patients do not respond to the treatment given, and this study has suggested alternative treatment strategies for further testing for these patients. In our interactome study we have described functions of miRNAs in relations to both mRNA and protein, and how the regulation in breast cancer is affected. This is important knowlegde in understanding the mechanisms of miRNAs in breast cancer, as well as to understand how the different genomic levels in breast cancer is regulated.