Abstract 2467: Heme-CMap: Generation and characterization of ~20K L1000 profiles across 11 hematologic malignant lines

Abstract

Abstract The Next Generation Connectivity Map (CMap) provides a resource to explore the relationships between diseases, genetic perturbations and compound treatments at the transcriptome level. Current CMap focuses on 9 tumor cell lines of various cancer types, but none of hematological malignancies. Considering the distinctive features of hematologic malignancies at both transcriptome and genomics alteration level, we recognized the pressing need for a similar resource of hematologic malignancies to better understand their unique characteristics of compound treatment response. Here we generated the “Heme-CMap” which included 7 lymphoid and 4 myeloid lines. In order to better understand the compound effects on hematological malignant lines, especially lineage specificity of lymphoid versus myeloid lines, we selected 100 compounds targeting B-cell lymphoma and 38 compounds targeting cancer pathways and hallmarks in our reference compound list. Each compound was tested at 4 time points and 6 doses. To this end, the Heme-CMap encompasses 20,472 compound perturbations. The size of this dataset and compound collection allowed us not only to globally compare the classic CMap to the newly generated Heme-CMap, but also to explore cell lineage effects of lymphoid versus myeloid cell lines within Heme-CMap. We found that several perturbation classes showed differential activity between classic CMap and Heme-CMap while some pathways show coherent activity. For example, EGFR inhibitors showed coherent activity in both classic CMap and Heme-CMap. On the other hand, BTK inhibitors showed coherent effects only in a subset of the lymphoid lines, but not the myeloid lines, consistent with the function in altering B cell signaling. We further investigated the compounds in Heme-CMap and observed samples clustered into two major groups by mechanism of action (MoA) of the compounds. One group contained samples treated with compounds targeting growth and survival, and B-cell signaling pathways, while the other group contained samples treated with compounds causing cell killing. Interestingly, we found both groups in the lymphoid lines but only the cell killing group in the myeloid lines. This may reflect our compound selection targeting signaling pathways implicated B cell lymphoma. These results suggest cell lineage plays an important role in compound response, and the experimental design of Heme-CMap allowed us to observe the differential compound effect in lymphoid and myeloid lines. We also found different types of compounds showed maximum effects at different timepoints and doses, suggesting various timepoints and doses can provide insights of compound MoA. In summary, our results demonstrate that Heme-CMap with lineage specific cell lines can elucidate compound MoAs complementary to classic CMap. Additionally, this effort also serves as a roadmap for generating data collaboratively with CMap on a large scale. Citation Format: Chia-Ling Huang, Andrew Yang, Ted Natoli, Lev Litichevskiy, Frederic Vaillancourt, Alan Rolfe, Yonghong Xiao, Aravind Subramanian, Lihua Yu. Heme-CMap: Generation and characterization of ~20K L1000 profiles across 11 hematologic malignant lines [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2467.