G5 Artikkeliväitöskirja
Identification of predictive ERBB mutations for targeted treatment
Tekijät: Koivu Marika
Kustantaja: University of Turku
Kustannuspaikka: Turku
Julkaisuvuosi: 2023
ISBN: 978-951-29-9285-0
eISBN: 978-951-29-9286-7
Verkko-osoite: https://urn.fi/URN:ISBN:978-951-29-9286-7
Predictive biomarkers, such as genetic alterations, are used in personalized cancer medicine to target treatment. Genes encoding members of the ERBB family of receptor tyrosine kinases are well known to harbor genetic aberrations that can drive cancer. These activating gene amplifications or mutations in the tyrosine kinase domain make these receptors potential targets for drugs, such as antibodies or tyrosine kinase inhibitors. Subsequently, several drugs targeting these receptors have been approved for clinical use. In an unselected patient population, the response rate to targeted therapy remains suboptimal, and the development of treatment resistance is essentially inevitable. Thus, there is a need to identify new biomarkers that predict drug responses more efficiently. However, the identification of new predictive mutations among the thousands of mutations discovered in patients requires resources. Several genomic studies and clinical trials have been carried out to provide functional information about the tumors. However, these efforts have highlighted the significant heterogeneity of tumors, and future work is required to take advantage of the full potential of these data.
The aim of this thesis was to screen for predictive ERBB mutations among the thousands of theoretically possible genetic alterations. This work presents the results of two screens with different setups. The first screen took advantage of publicly available cancer cell line databases that contain sequencing and drug response data for ERBB mutated cancer cell lines. For the second approach, the in vitro screen for activating mutations (iSCREAM) platform, previously developed in our laboratory, was modified to allow unbiased simultaneous analysis of thousands of activating mutations in ERBB3. Altogether, 79 potentially actionable ERBB mutations were identified. Detailed structural, biochemical, and functional analyses validated six of these mutations as novel activating ERBB variants with potential predictive value.
These results demonstrate that there are uncharacterized actionable ERBB mutations that can be identified with high-throughput screens. The mutations identified here were distributed across all four ERBB receptors and exhibited different gain-of-function mechanisms. The presence of multiple mutations in our screens also emphasizes the complexity of mutational profiles and that co-occurring mutations may promote additive functional effects.