Leveraging the Use of 3D Models to Identify and Target Novel Pathways for Colorectal Cancer Therapy

Abstract

A majority of human cancers are carcinomas, meaning they arise from the epithelium, highly organized tissues which serve as a barrier between the inside of the body and the outside world. Epitheliums are composed of a single layer of polarized epithelial cells which have biochemically distinct apical and basolateral domains maintained by junctional protein complexes. This apicobasal polarity is essential for the barrier function of the epithelium and loss of polarity is a hallmark of cancer. Loss of polarity in cancer is often accompanied by the epithelial-to-mesenchymal transition (EMT). The EMT phenotype also impacts diseases prognosis and therapy outcomes, particularly in colorectal cancer (CRC). Despite its importance, the mechanisms driving polarity loss and drugs that reverse it remain understudied, due, in part, to the complete failure of typical live/dead 2D high-throughput screens to capture colony morphology or the lack of robustness of 3D screens. To address this gap, we designed a high-throughput phenotypic screen using 3D type I collagen cultures of CRC cells to assess morphological changes in colonies and identified several FDA-approved drugs that re-epithelialize CRC colonies. Two of these drugs, azithromycin and linezolid, increased colony circularity, enhanced E-cadherin membrane localization, caused transcriptomic changes consistent with EMT downregulation, and elevated sensitivity to the chemotherapeutic, irinotecan. A retrospective analysis of patient data demonstrated that the use of azithromycin in patients undergoing treatment for CRC with irinotecan had improved the 5-year survival compared to the chemotherapy alone. Based on their bacterial target, we hypothesized that azithromycin and linezolid inhibit mitochondrial protein translation to drive re-epithelialization. Consistent with this, inhibitors upstream and downstream of mitochondrial translation caused similar morphological changes observed with antibiotics. Linezolid treatment also caused a decrease in the expression of electron transport chain subunits. This study demonstrates the successful use of a phenotypic 3D screen to identify novel treatment strategies to improve outcomes for individuals with CRC.