Abstract Marker-based isolation of Cancer Stem-Like Cells (CSLCs), which can display preferential resistance to standard-of-care chemotherapy, has proven challenging, due to their very small frequency in breast cancer tissue, thus impeding their characterization, as well as identification and targeting of mechanistic dependencies. To address this challenge, we leveraged a systems biology framework for the protein-activity based characterization of CSLC biology at the single cell level and for the identification of drugs that could reprogram them to a differentiated, chemotherapy-sensitive state. To enrich for CSLCs we performed flow cytometry-based sorting of cells dissociated from 7 metastatic breast cancer patients, both TNBC and HR+, using the EpCAM and CD49f markers, followed by RNA-seq profiling. Protein activity-based clustering, using the VIPER algorithm (Alvarez et al. Nat Genet 2017), identified complementary cell states associated with high activity of either established CSLC markers (BMI1, NOTCH1, etc.) or markers of differentiated cells (GATA3, KRT18, etc.). VIPER-inferred Master Regulator (MR) proteins, identified by comparing highest versus lowest stemness score cells in each patient, were virtually identical across patients and independent of hormone receptor status. Following MR validation by pooled, CRISPR/Cas9-mediated KO followed by single cell RNA-seq profiling (CROP-seq) in cell lines comprising both CLSC and differentiated BRCA cells, we used the OncoTreat algorithm (Alvarez et al. Nat Genet 2018) to identify small molecule compounds capable of inverting their activity (MR-inverter drugs), thus potentially inducing differentiation or cell death. For this purpose, we leveraged gene expression profiles of BT-20 cells—which significantly recapitulate the CSLC MR signature—following treatment with a repertoire of 91 clinically relevant drugs. Albendazole, the top drug identified by OncoTreat, depleted CSLC population (p = 2.0 × 10−4, by Fisher’s exact test, w.r.t. CSLCs to differentiated cells ratio) without noticeable cytotoxicity in the experiment of single cell-based assays in a TNBC PDX model, at 14 days after treatment. In contrast, paclitaxel, a commonly used chemotherapeutic, significantly increase the ratio of CSLC to differentiated cells (p = 2.6 × 10−4, by Fisher’s exact test), suggesting that alternating treatment with the two drugs may abrogate the tumor-initiating potential of paclitaxel-resistant cells, while also preventing uncontrolled tumor growth. This rationale of combination therapy was confirmed by preclinical study, where treatment with multiple cycles of albendazole and paclitaxel displayed significant synergy (p = 0.00869), compared to the corresponding monotherapies. In conclusion, our systems biology framework is a proof-of-concept that drugs can be effectively repurposed to enhance the therapeutic activity of anti-tumor agents used in conventional chemotherapy, and this method can be generalizable to address the challenge of other cancers with heterogeneity. Citation Format: Jeremy Worley, Heeju Noh, Daoqi You, Mikko M Turunen, Hongxu Ding, Evan Paull, Aaron T Griffin, Adina Grunn, Kristina Guillan, Erin C Bush, Samantha J Brosius, Prabhjot S Mundi, Peter Sims, Piero Dalerba, Filemon S Dela Cruz, Andrew L Kung, Andrea Califano. Identification and pharmacological targeting of treatment-resistant, stem-like breast cancer cells for combination therapy [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr PR01.
Read full abstract