No abstract

e14506 Background: Mutations in the ligand binding domain (LBD) of estrogen receptor α (ER) confer constitutive transcriptional activity and resistance to endocrine therapies in breast cancer patients. Accumulating clinical data suggest adverse outcome for patients harboring tumors expressing these mutations. We aimed to elucidate mechanisms conferring this aggressive phenotype. Methods: Cells constitutively expressing physiologic levels of ER harboring activating LBD mutations were generated and characterized for viability, invasiveness and tumor formation in vivo. Gene expression profile was studied using RNAseq. Metabolic properties were assessed using global metabolite screen and direct measurement of metabolic activity. Response to chemotherapies was assessed using viability assays. Results: Cells expressing mutated ER showed increased proliferation, migration and in vivo tumorogenicity compared to cells expressing the WT-ER, even in the presence of estrogen. Experiments in mice revealed a more aggressive phenotype of the Y537S mutant compared to D538G as well as unique tissue predilection. Thus, 538G-ER cells exhibited liver tropism while Y537-ER cells mainly metastasized to the lungs and lymph nodes. Importantly, both mutations conferred resistance to paclitaxel and doxorubicin. Further studies indicated association of the mutated ER with upregulation of genes involved in tumor cell metabolism. Indeed, a global metabolic screen revealed distinct metabolic profile for cells harboring activating mutations, including the ability to utilize glutamine as an alternative carbon source. Moreover, we observed unique metabolic activities enabling cells to thrive in urea-reach environment of the liver, and overcome doxorubicin-induced stress responses. Conclusions: Taken together, these data indicate estrogen-independent rewiring of breast cancer cell metabolism by LBD-activating mutations, enabling aggressivene clinical behavior and chemotherapy resistance. Importantly, These unique metabolic activities may serve as a potential vulnerability and aid in the development of novel treatment strategies to overcome endocrine resistance.

Colorectal cancer (CRC) is currently the fourth leading etiology of brain metastasis (BM). Yet, mechanisms supporting the formation of CRC BM are mostly unknown. In order to identify drivers that lead to tropism and adaptation of CRC cells to the brain environment, we analyzed an extensive genomic database, consisting of over 36,000 human CRC primary and metastasis samples. Several genomic alterations specific for BM were noted, among them increased prevalence of insulin receptor substrate 2 (IRS2) amplification, observed in 7.6% of BM, compared to only 2.9% of primary tumors or other metastatic sites (p<7E-05). This observation was validated by Immunohistochemistry studies of human clinical samples, showing increased expression of IRS2 protein in BM. IRS2 is a cytoplasmic adaptor mediating effects of insulin and IGF-1 receptors and is involved in more aggressive behavior of different cancer types. In order to study the ability of IRS2 to promote growth of CRC cells under brain microenvironment conditions, we employed an in vitro system consisting of cultured human astrocytes or their conditioned media. Indeed, IRS2-overexpressed CRC cells survived better and formed larger 3D spheres when grown in brain-mimicking conditions, while IRS2-silenced CRC cells showed a mirror image. Similarly, In vivo studies, using intracranial CRC BM mouse model, demonstrated that IRS2-overexpressed cells generated larger brain lesions, while silencing IRS2 dramatically decreased tumor outgrowth and extended survival. Transcriptomic analysis revealed enrichment of oxidative phosphorylation (OXPHOS) and Wnt/β-catenin pathways by IRS2. Indeed, IRS2-expressing cells showed increased mitochondrial activity and glycolysis-independent viability. Furthermore, IRS2-expressing cells had increased β-catenin transcriptional activity, and either β-catenin inhibition or IRS2 inhibition in IRS2-expressing cells decreased their viability, β-catenin transcriptional activity, and mitochondrial activity. These data suggest involvement of IRS2 in modulating OXPHOS through β-catenin. Exploiting this mechanism as a potential vulnerability allowed us to develop novel treatment strategies against CRC BM. NT219 is a novel IRS1/2 inhibitor already being tested in clinical trials. Treatment of mice harboring CRC BM with NT219 and 5-flourouracil reduced tumor growth and prolonged mice survival. These data reveal, for the first time, the unique genomic profile of CRC BM and imply the IRS2 role in promoting CRC BM. These effects may be mediated, at least in part, by modulation of the β-catenin and OXPHOS pathway. These findings may pave the way for clinical trials evaluation this novel IRS2-based strategy for the treatment of CRC BM.Accession no GSE203017 and here is the link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE203017